forgive me, forget you: villnoire in review

caution: this review contains just about every spoiler in villnoire.

villnoire's moralizing about forgiveness denies that victims of racist violence (or even just personal wrongdoing) have a legitimate right to their feelings about it. the wisest survivors sympathize with the complex situation of their offender, and excuse his actions without reservation or resentment. 

when villnoire does allow survivors to show pain, it's in the form of anger, which the game condemns as a deathtrap, devoid of any value, destined to consume them unless they purge it by forgiving the perpetrator. 

moreover, villnoire shows us time and again, anger is dangerous. it's a slippery slope to murderous retribution. and, as the party's moral authority and sole female voice insists, murder is always wrong.

when it comes time to face the big bad, the genocidal fascist king vorian, villnoire stresses the need to extend that principle of forgiveness to him, but this practice is absent from the ending. it's of the utmost importance that we forgive vorian, but somehow not quite important enough to show. forgiveness is only worth depicting when it's for the protagonists. why?

one of the most important observations villnoire makes is that the ultimate cause of racist violence isn't the individuals that perpetuate it, but the conditions that create them. vivian argues that this is why assassinating vorian isn't just wrong (another can of worms), it would be an ineffective solution.

so when, at the end of the game, the party enacts a coup d'etat, incarcerating the racist king so his kindhearted and non-racist brother takes the throne, the story betrays its own arguments.

the game's understanding of systemic racism is proven totally immaterial. to replace an evil autocrat with a benevolent one and call it a day suggests that the oppression of druids was the product of one man's cruelty after all - a point the game text denies repeatedly.

instead of meaningfully challenging the hereditary monarchy or societal attitudes that created and entrenched vorian, villnoire mostly uses the notion of systemic violence to absolve the protagonist of culpability for his participation in genocide.

in the end, the game only commits to its ideals of forgiveness when it means people stop being mad at lukas. he's a good person deep down. the only moral option is to forgive him. to villnoire, anger is a kind of harm - it's almost murder, after all - and good people don't deserve that.

what do bad ones deserve? if villnoire believed forgiveness was meant to for the worst of the worst, it would show it. the opportunity to have second chances, to learn and grow and do better, the right to prove your deep-down goodness is available on-screen only for those we already know to be fundamentally good. the fate of bad people doesn't matter enough to warrant concern or camera, as long as there's compassion for the good. that isn't forgiveness. that isn't justice, love, or revolution. that's just solipsism.

the hollowness of villnoire's theming is disappointing because the foundation of a better story is there. the first half of the game sets up a lot of interesting relationships and tensions between characters. it almost tells an affecting (if simplistic) story of found family among those harmed by fascism across all levels of the social hierarchy, and from start to finish it feels like a complete rpg narrative without ever dragging or becoming tedious. it even teases the possibility of a more radical politics of forgiveness and transformation. but the second half can't deliver on those promises in a satisfying way because the because story cares more about vindicating lukas than about empathizing with the victims of the atrocity committed through him.

there's much more worth critiquing in villnoire. i've hardly touched on its untenable stance on violence (revolutionary violence is necessary and good, as long as you don't kill anyone), or its subtle but pervasive sexism, or the many other ways it misunderstands systemic racism. despite these issues, i enjoyed my time with villnoire. i wanted it to resonate more than it did. that means i saw something good in it - so i have to forgive its faults, right? it didn't materialize in this game, but maybe it will in the next one. i look forward to seeing what little wing guy makes next.

you can download villnoire for free at itch.io or rpgmaker.net.

for more thoughts on the game, read dari's review at indie hell zone.

the bad fight

so i've spent the past 6 months working on and off on rpgmaker games. mostly on, to my credit, and a lot of the off was spent writing. but in my mental task list, the rpg i started formally working on in december is my main endeavor, and everything else has been a side project. through all of this i've been holding onto this idea of making short rpgmaker games, really simple, straightforward ones i could complete in a week or two, that are just series of battles without story or context, just playing with different mechanical ideas within the confines of rm2k's default battle system. party dynamics, enemy balance questions, feelings of battle. ironically these supposedly lightweight, non-blocking rpg experiments have been blocked in my mental task queue behind the main rpg i'm working on - which was supposed to be short and lightweight itself, before i stepped into the whirlpool of adding a "simple" framing story. but they've remained on the horizon as something to do after, even if i find myself too busy with work to pursue larger projects. (actually, i think this idea came about whenever it was i played 50 short games)

well, finding myself with some free time yesterday evening, too late in the day for a more concerted work effort, i started sketching out ideas for some of these simple 'battle series' games. trying to lay out an interesting 2-person party dynamic i encountered some familiar tensions.

there's a notable difference between designing encounters to be completed in a specific order by a fixed party, and making random encounters to be fought in arbitrary order and number. the first invites question: why this fight, this way, at this time? why are the fights in this order? what makes this experience, which has been individually curated, worthwhile? there's a pressure to make it internally satisfying; it invites a kind of puzzle-designing approach. what is the solution? random encounters on the other hand have an answer handy to all these questions: it was chance. you weren't "meant" to have this specific fight at this specific time, under these specific conditions. the rules were written to define a possibility space, and this specific cadence of fights emerged from that space by the forces of randomness. it's essentially generative, in a simple way. the goal in writing this isn't to put internal meaning on individual fights, but on their sum total. to overdesign random encounters with narrow paths to victory creates formal repetition, which can feel like a waste of time. to leave them loosely designed creates only incidental repetition, which can be filtered out. it's easy to be angry at a scripted series of fights for being one fight too long, failing to end at the perfect moment and becoming a chore. it's harder to be angry at a random sequence for the same. the rng doesn't know better. the designer does.

it's scary to underdesign a fixed sequence of fights because if the purpose of an individual fight isn't clear, it feels like filler. it has no point, it's just there to artificially extend the game. it's harder to trust that it has a long-term purpose. even if it whittles you down, why wasn't it at least fun? why curate individual challenges if you're not going to make them good? this is the logic that spins in my head. i don't necessarily agree with it, but it's hard not to be give into it, consciously or not. i'm doing level design, which is like game design for levels, and it is right to make your game design good game design. prescriptive notions of quality haunt every step of the process, largely because i don't have any other logic available to turn to when it comes time to create something from nothing. i've seen "good game design", so i can attempt to imitate it inside the constraints i've created for myself through the decisions i could make on my own.

i feel unsatisfied as i do this because i know that it's limiting. in practice it reduces individual battles to math puzzles with optimal solutions built in. i fell asleep last night thinking about possibilities for a single-character battle series. i wanted fighting to be fast and impress a feeling of power despite being alone, but i didn't want battles to be trivial or repetitive. in order to stand up against superior forces, i'll give the main character a multi-target stun. ah, but why wouldn't i use that EVERY battle, trivializing any decision making process? well, maybe it doesn't work against everything, now i'll introduce a second enemy, one that necessitates a different response. maybe it's a hawk and it needs to be bound with gravity magic to neutralize it. now when you face a mixed group of enemies there's a meaningful decision to be made: do i disable the hawk with gravity magic, or stun the rest of its allies? that's "good game design" baby. but these meaningful decisions add up to something i think is boring. the end product for the player is a sort of internal algorithm, a mental decision tree for arriving at the right strategy for a given circumstance. i've written about this already - this is the ethos of lufia and the fortress of doom, and of my current rpgmaker project, which is concerned with the same concept. i'm not knocking this approach - it has important advantages over the even more repetitive model where enemy groups offer no resistance and the party's "algorithm" is exceedingly simple, or fails to make use of the whole breadth of their skillset. but i think it leads to me feeling like i'm making the same game over and over again. i don't want to just make better and better algorithm-builders. i want to go beyond the feeling of having one core approach and stretching it into new situations. as many good experiences as i've had from the angle of "if this fight is a puzzle, how do i solve it?" (lufia 1, into the breach, radiant historia...), it's a design ethos that i think leads to the same set of experiences. what other experiences are possible with the language of turn-based rpg combat? i like rpg combat systems so much - but what does it give me, what can it give me, beyond the satisfaction of using a fun toolkit to great effect?

so i think i have two options for where to go from here. one is to keep leaning into the algorithmic strategy angle. i still haven't done it with non-fixed encounters, and i think it might work better there because without the fixed order i can see it feeling less like grasping for the right key for a lock. i would like to one day take on the challenge of designing combat for a longer game, with mix-and-match party selection, and more complex mechanics than rpgmaker 2000 supports, where i can play more with larger, more complex character designs and where every ability feels like it plays a role in the grand strategy. in this context the spectre of "good game design" will surface in other ways, but at least it won't result in the exhaustion pressure to curate every minor encounter to perfection.

the other option is to stick with fixed battle series but let go of the pressure to overdesign. what possibilities open up if i let battles not be fun or individually challenging? are there other core activities available besides optimization, or can rpg killing/survival mechanics only build experiences about killing and survival? many other feelings come up in playing an rpg that are just taken as chaff, a byproduct of bad design or else the cost of admission, swallow it and forget about it. but the feeling of (say) another unwanted battle getting in the way of what you wanted to do is almost universal in rpgs. the action of rpgs is all about clearing obstacles, which can be fun but is often a chore. much of rpg design has become about maximizing the fun of clearing obstacles through interesting systems, customization, etc... on the assumption that the fun part is good, and the choreful part is bad. good rpgs are good because of the fun parts, which design should maximize, and in spite of the choreful parts, which should be minimized by design and ignored by players inasmuch as possible. is there room for something else? where? this is where i'm held back by not being a greater consumer/imaginer/understander of art, i don't have much of a basis on which to envision other possibilities. all i know is that when i'm submitting to the dictums of game design i feel like i'm painting myself into a corner.

**************

afterthought 1: one place i can look for inspiration is 7th saga, a game that rarely let me feel powerful. when it did, it was only empowering me enough to feel safe, or to get by reliably. there was hardly one easy win in the entire game. the battles were about survival, it really leaned into that, and the non-combat elements of the game's design reinforced it. here mundane encounters felt weighty because they all really were existential threats that could wipe the party if things went wrong (and they so easily could), and your best defense was to take them very seriously. exhausting to play, but full of intention. but that was a much more fully developed context to explore these fights in. it would be hard to convey any of this in a context-free series of fixed encounters.

afterthought 2: these question apply particularly to mundane encounters. optimization isn't usually a concern in boss battles. these ARE meant to be highly curated, and have more room to make for memorable and rightly challenging puzzles than mundane encounters. i'm less interested in designing these. i feel like boss battles are well understood. as a matter of game design they tend to have clear goals, they're the highest stakes, highest tension test of the player's ability to speak the rpg's mechanical language. a lot of memorable rpg boss fights have been made already. if i have any interest in boss fights it's in making them exciting without using gimmicks or new mechanics, grounding them in the fundamental rules of the game. i love a good gimmick boss, and i think it's good and great to break the formal constraints of a system to reach higher emotional peaks and unexpected experiences, but at the same time it often feels like a lack of confidence in the game's basic toolset.

how to never miss in lufia and the fortress of doom

lufia and the fortress of doom has no random miss chance for weapon attacks and damage spells. aside from the odd cursed sword, offensive moves land with 100% accuracy. it's an enormous mercy for a game with long meandering dungeons, full of dead ends, worthless treasures, and encounter rates that are simply insensitive.

this is not to say that attacks always land. one of the most common complaints about the game is its handling of targeting. fortress of doom joins enemies of the same species into "groups"; you can choose which group to throw a single-target attack at, but who it hits within that group is up to chance. a little strange, but you can still plan around it, especially with smaller enemy groups. the real point of frustration is the possibility of wasting turns by overcommitting to attacking one group. if two or more characters target the same group, and all the monsters in that group have died by the time the last character executes their attack, the attacker will swing at the empty space where an enemy used to stand. reviewers didn't understand why characters wouldn't redirect their attack to a living target, like most turn-based games allow - or at least default then to defending, like golden sun. instead, carelessly assigning all characters to attack the same target often means watching a parade of futile slashes play out one at a time.

i can't blame anyone for having a bad time with lufia. there are lots of valid issues to take with it, like its truly miserable dungeon design, its multiplying fetch quests, and its low opinion of women. but its targeting system, and specifically its "committed" targeting, is actually one of its smartest features. far from a sentence to unavoidably burn turns whiffing, my experience of lufia's battle system was of encounters that began and ended quickly, in which my mindfulness was rewarded with the satisfaction of watching my plans play out just as i intended and without waste, despite the uncertainty faced. i would estimate the rate at i lost attacks to overcommitment to be less than 1 in 50 - fewer than i'd lose to random chance in most games - and this low rate was easily attainable without memorizing enemy health values, tracking damage inflicted, or going out of my way to learn how enemies work. even if my experience isn't representative, it's still possible, without exerting undue effort or relying on luck.

FoD's solution to overcommitment is in establishing implicit rules for ascertaining enemy durability. these rules are broad and have exceptions, but are generally maintained. they are especially pronounced in the game's second act; they stabilize after elfrea, once you have a full party and jerin has artea's bow; and they destabilize in the third act for reasons we'll explore later. the rules allowed me to parse battle scenarios quickly and determine the quickest and least wasteful way of distributing my attacks. because the rules were consistent, i could make these judgments even when fighting unfamiliar monsters. instead of having to figure out how each new enemy worked from scratch, i only had to learn the exceptions to the rules. this approach didn't just let me leverage my knowledge of the rules to assess enemy health and allocate attacks efficiently, it created an environment of exploitable normalcy where experience and intuition translated into mastery. as long as the rules applied, i was in control.

*

FoD's enemy evaluation rules are the player-facing effects of choices in the game's numerical balance. a precise and consistent relationship exists between the damage you deal as a player and the amount of health an enemy has. the nature of this relationship is to make it easy to gauge which, and how many, attacks should be committed to each enemy. it supports an understanding of enemy health in qualitative terms rather than quantitative ones. that is, decision-making happens less often on the level of "how many points of damage?" than the level of "what combination of specific actions?". although this is not a unique space for an rpg battle system to exist in, the move away from mental math and value tracking is essential for making overcommitment avoidable.

i identify 4 design choices that empowered me to make quick, clean judgments about targeting commitments, either by generating rules, broadening the applicability of rules, or limiting the amount of content to be learned and understood:

1. low upper health bound. mundane enemies (i.e. non-bosses) have low HP relative to the player's damage. any given monster will die to some combination of, at most, 2 character's standard attacks. most will even die in one, if it's the right one. fortress of doom's tight adherence to this upper bound teaches you over time that it's wasteful to commit any combination of 3 attacks, and certain combinations of 2 attacks, to one enemy up front. the rare exception exists - the trap harps near bakku come to mind, but because they're also huge, weird, and unfamiliar, it reads as a measured irregularity to factor into your strategy of rather than a contradiction of the whole paradigm. in general, two hits were the most it took to kill anything. (footnote one)

2. close lower health bound. as low as enemy health is, it's still high enough that weaker moves, such as jerin and lufia's weapon attacks, will almost never one-shot. this is a subtler decision, but it limits the number of situations in which a weak attack can substitute for a strong one, emphasizing the difference between them in strength and purpose. in fortress of doom, two characters' attacks are never interchangeable - it always matters who is attacking whom. (footnote 2)

3: consistent attack outcomes. a given attack tends to have the same qualitative outcome against all current targets, regardless of damage and health values. the proximity of the upper and lower health bounds means that this consistency will hold within an area, and continue to hold upon progressing to a new area, as enemy stats are balanced around maintaining these bounds. consistency means that a given attack will map reliably onto a qualitative outcome: aguro's basic attack is a one-hit kill, it will one-hit kill most enemies in an area, and although it may drift into two-hit kill territory for some enemies, you can rely on it dealing damage in one-shot range throughout the current area and into the next one. the same consistency can be expected of the hero's attack, which falls in the upper end of the 2-hit kill range for all current targets. the major exceptions lie in elemental damage spells, where weakness and resistance can be tricky to learn and exploit (the main caster's elements are lightning, water, and explosions, so...), and in weapon attacks against enemies like will-o-wisps with physical resistance. 

4: low number of relevant attacks. "hero" and aguro only have physical attacks. jerin has two sets of multi-target attack spells, but they're overshadowed by her faster, cheaper, and often stronger bow. it's lufia who has the widest range of spells to worry about, although her single-target lightning spells provide the clearest benchmark for comparison. the other three party members are strongly identified with their weapon attacks. the low number of attacks on hand means fewer outcome mappings to learn and remember. 

the cumulative result of these features was that i came to understood attacks in relation to each other. if every enemy can die in some combination of two attacks, then what should i pair each attack with in order to kill? thus did FoD let me avoid numerical logic in strategic play by learning what attacks pair with each other. the function, the identity of lufia's "bolt" spell is not in the value of the damage it does but the fact that, whatever the damage, it will finish off an enemy weakened by jerin's bow or the hero's sword. the function and identity of aguro's attack is that it needs no complement; that of lufia's weapon attack is that the only complement so strong that it could kill on its own. consistency of outcome means that established pairings remain lethal as you deploy them in new contexts; you could call each pairing a rule in itself.

these rules form the body of exploitable knowledge that transfers between areas; they simplify the process of getting acquainted with new enemies and limit learning to the exceptions. whatever comes next, experience shows, won't be so strong that it can survive jerin and the hero working together. when i go to a new area, the numbers go up a little bit, the rules remain true. i can still count on it to provide an upper limit for how much it'll take for me to kill an enemy. where rules fail to apply, it's never for long, and it errs on the side of enemies being stronger than expected (i have the "old cave" in mind here). never does lufia surprise you with an enemy that dies too quickly. these strong bounds on enemy health and consistent outcomes from attacks, in terms of qualitative actions, are what kept overcommitment from ever becoming an issue.

the carefully established and maintained rules formed the basis for engaging turn-based combat. at its best, fortress of doom uses limited exceptions to the rules, in combination with elemental weakness/resistance and random enemy group composition, to keep encounters varied. for my tastes, it strikes a good balance between keeping encounters varied enough to avoid feeling repetitive, but straightforward enough to resist fatigue of constant strategizing. it asks just enough to keep me mindful through the high volume of battles and rewards that mindfulness by ending battles faster. unfortunately, FoD is not always at its best, and neglects to push back hard or regularly enough, falling into tedium and repetition only exacerbated by its ever creeping encounter rate and compounding fetch quests. but its high points are high and many enough to illuminate the role of player knowledge in rpg combat experiences, and to demonstrate the value of centering knowledge in enemy and encounter design through rules and exceptions. i hope that by describing the logic of FoD's combat design in detail i can show how many ideas it boasted worthy of new adoption and further exploration.

*

as long as the rules applied, i was in control. what happened when they didn't apply anymore?

toward the end of the game, the rules cease to apply. during the introduction to the game, you play as maxim and the heroes of the first doom island war, navigating a dungeon full of encounters against hydras and towering efreeti that look dangerous, but are trivial for your party of level 75+ veterans. at the start of the third act, when the new generation of heroes arrives on the epro continent after uncovering dual blade, many of the enemies seen in the introduction begin to appear again. symbolically this represents the return of doom island in full force: you're fighting the sinistrals for real now, just like maxim did.

the increase in stakes is accompanied by a difficulty spike. not only do you face largely unfamiliar enemies, none of whom carry over from previous areas or even, really, resemble monsters you've seen before, but also the the low hp rule suddenly falters: enemy health jumps up and previously trusted lethal pairings are no longer sufficient to kill. the relationships between party members are further upset by the arrival of "high defense" enemies like mega turtles that take drastically less damage from jerin, preventing her from "complementing" the hero or even aguro's attacks. this gets worse and never gets better. enemies hit harder and harder and by the final dungeon represent a random encounter can represent an existential threat even at full health. and right before the final dungeon, lufia departs from your party, leaving you to face these hardest fights without her, plus three of the sinistrals, returning only for the final final battle against guard daos. (hope you remembered to equip the thunder ring before nazeby!) (footnote 3)

the cumulative effect here is that the vanishing rules (footnote 4) erase the ability you spent act 2 mastering to judge enemies, allocate attacks efficiently, and control the flow of battle. instead you spend the endgame grappling with vulnerability in increasingly hostile circumstances. fortress of doom isn't interested in the experience of powerlessness per se (like, say, live-a-live), but it does use these feelings to underscore the emotions of the whole ending. lufia sees there's something powerful in pitting the player against the final boss when they're at their most desperate, rather than when they're feeling their strongest (like they might after a series of ultimate weapon sidequests). the conflict of lufia 1's ending extends past the the final battle as (spoilers again) lufia demands the hero kill her to stop the sinistrals' cycle of resurrection. dual blade rids her of erim's spirit, but as she lays wounded from the schism, the floor crumbles beneath her and she disappears into the abyss. the game lets us wallow in the hero's shock at his best friend and beloved's abrupt disappearance, and when the party finally escapes, he leaves alone without celebrating or saying goodbye. it's genuinely a very affecting sequence. it denies you an opportunity to celebrate a hard-earned victory after a harrowing final dungeon (with no save points, incidentally). the precarity of the final dungeon is instrumental in setting the tone that drives home the tragedy of the ensuing events. i worked so hard to get here, and now, for what? i was already at my limit.

hence the careful balancing and disruption of enemy health and player damage doesn't just create appropriate difficulty in line with the escalation of challenge as the game progresses, but informs the emotional arc of the whole third act as well. fortress of doom reminds us that simple systems are perfectly capable of supporting mindful and rewarding rpg combat, that they're perfectly capable of coloring and augmenting a story, and that rpg developers have been thinking about the connection between gameplay and narrative for a very long time. as certain experiences and aesthetics have been codified as "good game design" in the years since lufia, it's easy to misunderstand the values that consciously informed fortress of doom's conscious design (footnote 5). but all the signs point to its core design being thoughtful, deliberate, and effective in delivering a specific emotional experience.

----------------------

footnote 1: this upper bound assumes you're current on levels and equipment and fighting current enemies, although when the encounter rate is so high and sweet waters cost 8 gold apiece, i wonder which level is ever "current". for my part, i never intentionally grinded, but i also i rarely ran from battles or used sweet waters to lower the encounter rate, because i liked fighting them. anyway, the same assumptions are in place for all my generalizations

footnote 2: when we're dealing with enemies at full health, anyway. naturally, if an enemy is on its last hit point, it doesn't matter who swings their sword.

footnote 3: writing about rpg mechanics is hard because it's hard to tell whether your experience was the norm, or relegated to your particular way of playing the game. it was during the third act that i stopped fighting every battle and started chugging sweet waters and lobbing smoke balls. did the difficulty surge because i fell behind in levels? or was this my return to baseline after 30 hours spent overleveled?

footnote 4: i suppose another rule in the norm could put an upper limit on how much damage enemies do to you. for instance, i never saw an enemy who could deal as much as half my health in a single attack. the biggest dangers were groups of red magi and other enemies who could hit my whole party for a quarter to a third of their health, multiple times in a row - and even those felt like outliers. i think they all vanished by elfrea

footnote 5: even lufia 2, published less than 2 years later, discarded much of fortress of doom's innovation in favor of ideas that were visibly more "final fantasy" or "zelda". incidentally, of the two titles, it's the one more widely celebrated by retro rpg fans

new decade, new name

gay greetings and happy 2020! to mark the new decade, i'm trying on a name that better reflects my evolving relationship to gender and my ambiguous position in racial space. in 2019 i published some writing under the name "Sraëka Lillian". i've been trialing it in limited slices of my personal life too, and having sat with it a few months now, i'm ready to debut it for public usage. old names remain viable: you don't have to avoid calling me clayton or any other name we have history under. but in the spirit of the new decade, i invite you to try using Sraëka as your primary address for me instead, online and off.

pronunciation: SRAH-eh-ka. three syllables, emphasis on the first. the two dots on the ë are a diaeresis and indicate that a vowel is pronounced separately from the one before it. i'm happy to help if you ask. in a pinch, Lillian is fine too. it's less a surname than a backup.

pronouns remain the same as ever: inconclusive. none of the popular options are wrong, but none is quite right, either. follow your intuitions! be playful! alternatively, neutral language is a safe bet.

the temporality of support conversations in three houses

i succumbed: i didn't expect to love the characters of fire emblem: 3 houses so much. after the 3ds fire emblems i wasn't sure the series had anything left to offer me, but the students in black eagle house have won me over, and despite my grievances with the game in general, i'm still looking forward to playing the other routes just to get to know their characters better. their success is that they're fun and charming and well-acted and robust enough this time to justify the amount of time they spend in the spotlight. but most of their characterization happens in support conversations, which have a troubled structural relationship to the rest of the game. i don't feel this relationship has been very well considered. rather i feel like three houses has only heaped more weight on a structure that already undermined the potency of the stories the series uses it to tell

i'm still playing my first route but i wanted to record my feelings while they're loud in my mind. expect general spoilers up through the first couple post-timeskip chapters of the black eagles route


1. the sequestering of most character development in optional support conversations means that no character's growth or relationships are reflected in the main mission cutscenes. instead, when plot-inessential characters (ie everyone besides edelgard and hubert) appears in the main story, it's in the context of everyone in the house taking turns saying how they feel about the situation based on their one stable character trait - linhardt's sleepy, petra's not from here, and so on. i appreciate that some effort has been made to make characters respond directly to each other in these scenes, but such replies are the exception to the rule, and are still agnostic of whatever shared experience the interlocutors have been shown to have in their support conversations. the result for me is that, immediately, these 8-person conversations feel forced and formulaic, and more widely, so does the whole premise of the black eagle strike force. i'm not sold on anyone's reason for being here. whatever selfhood they're shown to have, it feels like their main reason for sticking around, for being generals in this war that didn't go anywhere until i showed up, is so that i (that's me!) can use them for video game stuff

the restriction harms in both directions. non-major characters are discouraged from growing in direct response to plot events as these developments can't be reflected in support conversations that were accessible before those events. the black eagles' lukewarm reaction circles are about the most they can affect each other without creating feelings that have to be addressed or reflected in the support side.


2. the flexible timing of support conversations means that they aren't guaranteed to occur at a time that makes any sense narratively or emotionally. in an effort to mitigate this many are locked until (or after) certain story events have happened, but not necessarily because the supports refer textually or emotionally to those events. the weirdest example of this is that my b-rank with leonie was locked until jeralt died, after which i was prompted to view it almost immediatey. but the conversation made no reference to her mentor's death whatsoever. in fact, it almost read as a bitter taste of dramatic irony, except that the tension never resolved. it was just a weird angry exchange about my not appreciating jeralt enough. oh, my god.. has no one told her yet? in the following explore phase, i found leonie gloomy with grief but seemingly completely oblivious of our recent conversation. it's possible to read this as an intentional writing choice, if a clunky one, but mostly it feels like an unresolved problem proceeding from a desire to make both these conversations available to you in any order

other common disconnects include: supports locked until after the timeskip that would have made at least as much sense beforehand, or that refer to the events of previous supports as if they had happened last week and not 5 years ago (caspar and ashe a-rank comes to mind here); and c-rank conversations potentially available before the timeskip but not accessed until after. in addition to discordantly preserving the goofy high school energy of the first half, these scenes' continued availability limits the degree of change characters (much less their relationships) can undergo during the five year gap, lest these late-viewed c-ranks paint an inconsistent character picture. the effect in all these cases is a kind of chronological incoherence - compounded with contrivance of basing the second half of the game in garreg mach again (who's grading these certifications now, by the way), it feels like there's no distinction between the past and present, between the black eagles as starry-eyed students and as hardened war generals. it turns out the (honestly moving!) sadness edelgard expresses in edge of dawn was unfounded - high school never has to end after all!


3. the nonlinear and nonexhaustive consumption of support conversations means that basic character information may be repeated frequently because you might miss it if it's only available once. how many times has shamir explained that she's from dagda, and she's a mercenary, and she joined the knights because she owed rhea a debt, but that's all settled now so actually it's totally whatever that she's here with you? this is a particularly egregious case; in others, the game is willing to hide unique lore and character background (i'm thinking of alois and leonie's supports) and i'm more allured by the prospect of seeing a differently incomplete picture of each character in each run


4. when you come back from the timeskip, all your old teammates stand in a circle again and tell you how much they missed you, while showing a glimpse of how they've changed since you last saw them 5 years ago. sort of. the best reintroduction you get is actually in the explore phase, when you can talk to each character one on one. but the explore phase is optional, and even if you take it at the earliest possibility, a week of time passes first, during which time prompts may come up for any of byleth's supports that had their requirements met before the timeskip but were plot-locked until after. after that, you're given a chance to view any other unlocked support conversations through the menu (and if you were serious about raising support levels, you'll have a lot of them unlocked). but these supports are, crucially, not intended as reintroductions to these characters. they're written as though reintroductions have happened, or rather as though they aren't necessary as all (see item 2). the existing framework of plot-focused main story cutscenes and character-focused optional support conversations and explore phases doesn't make room for character reestablishment that isn't optional to view or prone to being mistimed.


fire emblem's vision is of many complex, well-rounded characters having lots of characterization and lots of unique relationships with each other. the way they've chosen of making character content bountiful but not overwhelming in having sequences of one-on-one conversations between characters and making them optional. the amount of dialogue written for three houses is staggering. it's inevitable compromises have to be made in order to reconcile all these different pieces with each other and with the main story. and clearly the narrative team is to some degree aware of the challenges of this approach as new efforts have been made in 3 houses to restrict these conversations in time and tie them to events in the main story. i can understand and respect the task before them - it's an achievement what they've made work.

but in choosing what compromises to make, what values have been prioritized? even more than past titles three houses is designed to indulge player control of their experience. support conversations by their nature are about providing narrative rewards that are responsive to the player's choices about which characters to use and which to pair up, and their expansion directs energy and emphasis away from the main story's fixed cutscenes (which feel really slim in 3h). meanwhile additions like the explore phase only add more pieces for the player to opt into or out of or to reorder - to say nothing of grander choices like which house to teach or whether to join or oppose edelgard. all the unsynchronized story avenues are a pleasure individually but when added up they muddy the trajectory of character development, bloat it and disrupt its pacing

three houses best resists these issues when it anchors scenes to points in narrative time and says, you can't view this anymore, it doesn't make any sense. the aforementioned leonie b support could so easily have been a home run if it had committed to being a direct reaction to jeralt's death. instead three houses buckles to a fear of denying content to the player and affords them as much control of what events are viewed and when and in what order, neglecting to consider how the structures that afford this control affect storytelling in favor of the brute force addition of even more content to address combinatorial possibilities