Grant Kuning set himself an impossible task when creating Sethian. Text parser games are already notorious for their inability to handle the full range of sentences players throw their way. One that operates entirely in an unfamiliar alien language is doomed from the start to magnify the inherent obtuseness of the form.

So it is that Sethian teeters between intense hand-holding, telling the player more or less exactly what to enter; and absolute inscrutability, with no clear way forward while the AI rejects even the most reasonable and well-formed sentences. I got the bad ending just by fucking around to see if I could find any sentence the game would recognize, and the good ending only by looking up a walkthrough online.

But despite its inevitable failure, it's hard not to be a bit intoxicated by the sheer ambition of what Sethian attempts to be. Where other games may incorporate decoding a language as a sidequest or step in a puzzle chain, Sethian makes it the heart of the experience. At its best comes close to the excitement of playing Her Story for the first time, feeling your way through a world of possible queries, the only boundary being your own limited vocabulary. It presents a challenge to designers who come after: what can you grow from the seeds that Sethian planted?

I've never been good at picking up new languages. I struggle to memorize vocabulary, and conjugations fly right out of my head the moment it's time to form a sentence. Despite spending my entire schooling in various different language courses, I never picked up enough to be even conversant.

This is a source of great dismay to me. I love talking to people, I love reading, and I want to approach other cultures on their own terms rather than asking them to switch to mine. So four years ago I decided to start learning just a little bit of French every day with the help of one of those language apps. Even after COVID swooped in to severely limit my spare energy, I made sure to at least practice a bit every day to avoid losing what progress I'd made.

Dépanneur Nocturne is a bilingual game to a degree that I've never seen a game attempt before. It's not a "choose you language in the settings menu" thing—it's a "step into the dépanneur and the clerk starts speaking French to you" thing. You can respond in French or in English, and that's what she'll use to speak to you unless you ask her to switch. It's exactly like walking into a store in a Francophone region.

Despite four years of fucking around with a language-learning app, I've never had the courage to try actually engaging with Francophone media in its native language. I didn't think I could hack it. Even with frequent dictionary consults, I thought the process would be too slow and painful to actually be engaging.

I would never have gone into a menu and selected "French" for this game. But sure, I can say bonsoir when I walk in the door, just like I'd do in real life. And I found I could mostly make sense of the ensuing conversation. And so it was that I made it through the entire game without ever asking Eugénie to switch to English.

Of course, I had to consult the dictionary plenty of times. But not so much that it kept me from appreciating the cozy atmosphere, the mysterious worldbuilding, and even the charming writing of this game. And through all of that was woven the warmth of the implicit invitation the game provided to play it in its own tongue without judgment or expectation. And that means the world to me.

Good clean fun. Metagoofs upon hyperjapes, just like the original but more. As dunks on the concept of sequels go it's no Metal Gear Solid 2, but it's plenty funny. My only real critique is that this doesn't expand at all on the sense of exploring the possibility space of the game—the whole thing is just reiterating the possibility space of the original and seeing what's different.

When I first played this just after it launched, I was blown away by how gracefully it interwove narrative and mechanical interests to inspire a feeling of wonder and infinite possibility. In the eight years since then, I've often compared it to other "modular narrative" games that also try use mechanics to allow players to guide themselves through story, and almost every time found 80 Days to come out on top.

Eight years is a long time, though, and I began to wonder if I might be remembering this game as more magnificent than it actually was—and thereby holding other games to not just a high standard, but an impossible one. So I decided to boot it up again on its 8th anniversary and see how well it held up.

Dear readers, it held up well. It really is true that almost nothing I've played since builds a self-guided narrative so effectively (with Sunless Skies being one notable exception). The formula itself isn't very complex: have many places that the player can traverse non-linearly; give each place (and each route between) them its own little micronarrative; create occasional connections between these narratives; and tie all of these into a simple resource system. But this simple formula gives rise to an incredible sense that every run is its own unique story told in tandem by the player and the game.

In my first Citizen Sleeper review, I expressed a hope that another game could be build on the strong mechanical foundation that Citizen Sleeper created but didn't quite maximize. DLC would seem like a great way to do that, tying a new story into Citizen Sleeper's world. Tragically, the DLC we got fails to meet my hopes in so many ways.

To begin, the choice to release a series of self-contained "episodes" plays directly against the strength of the form. Modular narrative games shine because they let the player explore a possibility space of story and construct their own understanding of the game-world out of the path they find through that space. Citizen Sleeper's base game already struggles to take advantage of this, with little variation in the way stories play out and no connection between those stories, but the DLC is much worse. It's so small that it's forced into linearity. There's simply not enough room for multiple meaningful paths.

Since it can't use its mechanics to foster a sense of modularity and exploration, it instead uses them for raw difficulty. It expects to be played in a late-game save with easy access to cash, food, and repairs, so mostly what this means is making a lot of high-difficulty dice rolls and obtaining a bunch of one particular item. Unfortunately, it's entirely possible to beat the base game without having a source of that item in the quantity demanded, so the wrong build (mine!) can get locked out of the best ending. It can still be completed, mercifully, but it's a totally unnecessary dagger of frustration.

Since this is notionally a narrative-focused game, I should mention: the plot is fine. It's a bit more simplistic than the chewier parts of the base game, and it suffers greatly from being unable to include any of the characters you likely befriended, but if you think of it as another quest alongside all those in the base game it's solidly in the middle of the pack. But this isn't a novel, or a collection of short stories; it's a game, and as such I expect it to do something interesting with its mechanics.

(And no, the copy editing did not improve.)

An unusual sort of tactics game where fog of war rules the day. Success in battle, especially the first time you encounter a scenario, hinges on accurately predicting your enemy's loadout and position. I suspect that the winning strategy for must fights is to hold back, fortify a perimeter, draw out units into vulnerable positions, and pick them off until you have overwhelming ground superiority.

Most units can't move and attack in the same turn, and even those that can will deal more damage if they're stationary. This, along with strict line-of-sight rules, gives the defender a substantial advantage: if the enemy wants to approach you, you get the first (or hardest) attack. It leads to a very slow and deliberate style of play that I can imagine someone really liking—but that person is not me.

Given how much unknown information is internet in the core design, it's an interesting decision to also add randomness to most attacks. While standard infantry attacks only do 1 damage, most other machine gun fire does 1-2 and anti-vehicle attacks tend to do 2-6—a colossal amount of variance. With the game already encouraging such a conservative approach, extra randomness just serves as drag, forcing the player to assume the worst will happen and slowing down already sluggish fights.

Evertried is a by-the-books turn-based tactics roguelite with one distinguishing conceit: a "focus" timer that goes down in real time if you don't land hits on enemies long enough. And I'll give it credit for elegance: it uses the tension between planning out moves and building focus cleverly, with one upgrade track that gives passive bonuses when you have high focus and another that provides tools which require time and planning to activate. It's a very designerly construction, careful interconnections of mechanics all laid out just so.

Unfortunately, it's just not that fun. The focus timer adds stress and disincentivizes experimenting with activated effects. Choosing to ignore focus puts you at a clear disadvantage with half your upgrade options being useless. The de facto strategy ends up being "die to enemies enough that you mostly have muscle memory for how to navigate their attacks quickly" when what I really want a tactics game to test is my planning and problem-solving skills.

Despite the structural similarities, this is a fundamentally different kind of game than a modern Zelda. From the first glance at the manual, it's clear that this is concerned first and foremost with secrets. The game is crammed chock full of hidden passages and actions, and past a certain point simply can't be progressed without learning to look in every nook and cranny. Even the manual itself slyly hides crucial information in plain sight.

Later Zeldas translated this preoccupation into a focus on puzzles, which has a totally different feeling. Where those games feel like going through the motions, performing a dance choreographed by the designer, Zelda 1 encourages a sense that any wall might hide an incredible treasure and in doing so makes the entire world feel magical. There's a risk in doing so that the game locks itself off from too much of its audience, but I at least was able to get through it with nary a glance at the wiki (and only gentle hinting as to the location of one optional item).

This game's other major interest, much more to its detriment, is combat. Because this game is less interested in puzzles, the dungeon design is structured much more like a gauntlet of combat challenges than anything else, and this gets old fast. Towards the end of the game, each room feels like it could just as easily have been procedurally generated, with a handful of enemies randomly selected from among Bubbles, Like-Likes, Darknuts (fuck those guys), and Wizrobes (FUCK THOSE GUYS)... occasionally with fireball-spewing statues thrown in for good measure.

The end result is a combat system that careens between trivial (most of the overworld sections) and nigh-impossible with very few points where it actually feels both relevant and fun. Making matters worse, most of the really heinous enemies are immune to the late-game weapon treasures, making those treasures largely irrelevant in turn. I've never been a big fan of Zelda combat in later iterations, and I guess they come by that shortcoming honestly.

Playing Neon White casually when I have several friends who consistently rank in the top 50 players in the world is very strange. My personal impressions are clouded more than a little by the secondhand excitement of watching someone grind out better and better times, cheering for them when they overtake their rival, laughing as they look at the excellent names on the leaderboard. I love Neon White the milieu and I've seen up close and personal why someone would adore this game.

I don't adore this game.

There were moments when I thought I might. In the early-middle chapters, the game achieved a graceful balance of complexity and precision that had me excited. The levels were bite-sized without being simple, the weapons were cleverly placed to introduce possibility without breaking the game, and I found myself improving my route run after run until I found something close to the best possible (even if my execution could never measure up to my friends').

But this joy was short-lived. New weapons were introduced that ran against the grain of the game, doubling down on onerous precision and execution requirements in a game that by its nature has plenty of both already. Optimal routing started requiring me to comb the screen for pixel-perfect shots, or (as I actually did for the last chapter) simply look them up online.

Even in its failures, there's a compellingly homemade quality about Neon White. It has such an excellent core idea that it's easy to forgive the various flawed design elements. And although Ben Esposito is a veteran game designer, this project requires a different kind of design than anything he's published previously, with crucial elements of play unfolding from tiny decisions—how many frames should this take? how much ammunition should that have? how will this type of gun inform the level design inform the player's experience?

I spent the latter third of this game waiting impatiently for it to either end or get better, so I doubt I'll pick it up again even if DLC drops. But I'm glad it exists, and I'm glad I played it through, if only so I'll have the context to fully appreciate watching my friends shave milliseconds off their personal bests.

Diet caffeine-free Symphony of the Night. The combat is quite solid, and the bosses that really leaned in on the element-shifting aspect of it were a lot of fun. Everything else was just porridge-bland ultra-linear Castlevania-alike.

Rarely have I played a game that's made me more intensely question why it's a game. I was attracted to Genesis Noir by its visual style, and that turns out to be about the only thing this game has going for it... so why not just make it a cartoon instead? The interactivity in the game ranges from "press forward to progress the plot" to "click arbitrarily around the screen until you figure out what obscure thing you're supposed to do to progress the plot", and never once does it justify having interactive components at all.

Fortunately for those who might feel obligated to play it anyway, the plot does nothing to make up for the mechanical unsophistication. The jumble of biblical and scientific metaphors are too arbitrary to have anything interesting to say and too obvious to be intriguing. They act like a mountain of unflavored buttercream frosting, hiding the fact that the paint-by-numbers noir plot underneath has barely anything going on at all.

It's a testament to how much I like Citizen Sleeper that I'm giving it four stars despite having the absolute worst copy-editing I've ever seen in a professionally-released video game. It is strewn with typos, spelling mistakes, grammatical errors, and a thoroughgoing lack of understanding of the typographical conventions of written English. It is embarrassing, particularly in a game whose focal point is its writing.

All that said... it's still a good game. It walks in the footsteps of similar "modular narrative" games like 80 Days and Sunless Skies in using a light resource-management system to drive the player to allocate their time between many different bite-sized threads of plot, creating an individual route through the game's story. This is one of my absolute favorite microgenres but a difficult one to get right, and I'm thrilled with any game that can pull it off competently.

Citizen Sleeper isn't as successful at this as its forebears, though. By the end of the game you have enough resources to do almost everything the game has on offer, and there turn out to be minimal ways that plots can differ from playthrough to playthrough. Starting a new save feels more onerous than exciting, knowing that so much of it will be clicking through the same dialog rather than forging new paths.

I often finish modular narrative games with a sense that the system on which its built still has untapped potential. The calculus of quests taking inputs and producing outputs while clocks tick down cycle after cycle is simple but expressive—it could give rise to such an intricate web of interlocking threads. Is it too much to hope for that, with the groundwork now laid down, another game could be created on top of it... this time with an editing pass?

"What if Elder Scrolls games only had quests and no combat" is a pretty compelling question and this game is a pretty compelling answer.

Christa insisted that I play this and then I beat her high score by thirteen thousand points

This review contains spoilers

High highs and low lows. The 3D art is bland, possibly because I have no nostalgia for the PS1 look and certainly because I have no love for the muddy desaturated color palette. On the other hand, the 2D art is as strong as it was in Anodyne 1, and indeed is made stronger by the benefit of multiple worlds each with a distinctive and appealing look.

The plot is achingly predictable--the god-analogue is revealed to be a cold ideologue who's strangling free will? Quelle surprise! But the smaller-scale character interactions are fun and the writing is mostly quite solid with a few great moments and fewer terrible ones.

I think the most frustrating thing about this is that its mechanical ambition extends beyond the capacity of its control scheme. Movement in both 3D and 2D sections of the game is clunky in a way that's surely an intentional riff on the consoles the game evokes, but conflicts painfully with the challenging maneuvers it occasionally demands. It's a question of suiting the design of the tasks you set before the player to the context in which the player acts.