Returnal somehow still stands in 2023 as one of very few genuinely current-gen-native video games, and you can tell. The graphics are slick, full of shaders, particles, and lighting that underscores its otherworldly vibe. It takes full advantage of the DualSense's haptics to make the game as tactile as it is visual—the half-trigger alt fire mode is particularly impressive in its fluidity of adding what's effectively an extra virtual button to the controller.

My problem is simply that I don't like shooters all that much. All the genuinely challenging fights in this game put a lot of weight onto the player's skill at shooting while dodging and weaving, and I'm neither great at that nor particularly interested in it. In some games, the roguelike design would help compensate for this by giving me the opportunity to assemble a synergy build that can simply overpower enemies despite my middling skill, but that's not Returnal: the build customization here doesn't have the multiplicative scaling necessary to really tap into that mode of play.

So I made it to the final boss of the second biome, died with the last shred of its health remaining, and will probably never return to this game.

I'll start with praise: Wo Long's spirit system is genuinely brilliant, elegantly intertwining Bloodborne's reward for aggression with Sekiro's posture- and deflect-oriented combat, giving an extra dimension to combat and organically rewarding the player for mixing and matching between normal attacks, heavy attacks, and martial arts. I think you could build an excellent game on top of that system, but I don't think Wo Long quite hits that mark.

The first chunk of this game, through about the Ayoe boss fight, is paced so that this combat system shines. But as time goes on, the difficulty ramps up in ways that specifically undermine its greatest strengths. Enemies start having long, frenetic normal-attack combos that are very difficult for mere mortals to consistently deflect and which simultaneously leave no room for normal attacks. Fights start to boil down to dodging out of range, memorizing the timing of exploitable critical attacks, and nothing more.

This is compounded as the game starts dumping multi-foe fights on you, demanding that you react to layered overlapping combos that are barely feasible on their own. It's not even the difficulty that's frustrating—you can smash your way through these fights with enough patience—but all the nuance that's compelling about the combat is gone. After beating Lu Bu and feeling nothing more than vague annoyance, I think I'm done with this game for good.

I've grabbed more than a handful of deckbuilding roguelikes off the app store shelves just to see what's up, and none has come close to impressing me as much as Dawncaster. While this game has its rough edges in places (balance in particular), in terms of raw fun it deserves to be in the conversation with titans of the genre like Slay the Spire and Monster Train. Rather than having strictly separate classes, it incorporates a Magic-style color system where your player levels up stats which contribute colored energy which in turn determines which cards are playable.

This nails the feeling, core to the genre, of gradually and carefully assembling an absolutely unbeatable synergy engine, taking carefully chosen risks to pay off later down the line. Each class has plenty of juicy themes and strategies which expand even further once you enable additional card sets. Cross-class synergies dovetail nicely with multiclass-specific cards, giving tons of depth and replayability. I was really shocked by how fun this was, and I'll definitely be returning to it again over time.

I think I have about two chapter of this in me every few years. This is my favorite Zachtronic puzzle game in a lot of ways, I think the core mechanics are very cool and the look of it is spectacular, but it's so so easy for me to get frustrated by the more complex puzzles. It takes so much effort to just figure out how to get atoms in place to be bonded properly it makes me want to curl up and take a nap. I salute people who will go all in on optimizing even the toughest puzzles, but I need to take another long break before tackling the last two chapters of the campaign.

I've played a fair handful of games like this, low-budget 3D Unity adventure games with minimal mechanical depth and lots of writing, and almost to a one they end up feeling kind of sloggy. AAfACRbD is the exception. The writing hits the perfect balance of terseness and humor to propel me through the game, the stage design is fun to explore and generally not overlarge, and there are even nice little mechanical touches like items that can boost your speed to avoid long slow movement sections.

The plot is also shockingly beautiful and more than a little heartwrenching. I cried multiple times, which is absolutely not what I expected from the first hour of the game. Like the rest of the writing, the relationship between the protagonist and his fiancee does a ton of heavy lifting with very little actual text.

This game just oozes cool, from the concept of a stealth infiltration game focused on pilfering everything you can get your hands on, to the bizarre but incredible droning score, to the moody and engaging cutscenes. So it's a damn shame that playing it feels so consistently like a slog. The level design tries to combine complex maze-y architecture with ahead-of-its-time light- and sound-based stealth, and while each half of this is fantastic in and of itself, they intersect in a really unfortunate way.

The experience of actually navigating a level in Thief is typically one of scrambling wildly through twisty little passages all alike, trying desperately to find your goal and collect enough treasure as the maps you're given become increasingly more vague. At the same time, these areas are peppered with enemies who outmatch you in combat and will hear you coming from meters away. Many is the time you'll turn a corner and find yourself face to face with an angry guard, forced to reload again and again as you try to get your bearings and find a way to get behind him with your blackjack.

The game is also studded with levels that aren't really about stealth at all, where the primary enemies are zombie who'll attack you when you get close no matter how quiet you are, devolving into a kind of straightforward dungeon crawler that doesn't play to the strengths of the engine. If I wanted to play King's Field, I'd just play it! All of which is to say: after making my way through the first seven missions (one of which took an hour and a half not counting reloads), I found that my reticence to slog through a bunch of zombies in The Haunted Cathedral outweighed my desire to see all the cool stuff left in the game.

The fundamentals of Mahōkenshi are great. Combining the Slay the Spire-esque roguelike deckbuilder with a tiled tactics board is a tall order, but this game nails it. Using energy for movement, adding landscape features that affect that movement, and giving each class access to different ways of subverting that movement economy through their card pool is brilliant. Giving enemies alert and chase radiuses so that they can guard treasures and give the player a choice between fighting, darting in and out, or avoiding the risk altogether is inspired. Bonus challenges recontextualize each mission and provide replay value even in the absence of a true procedural-generation mode. You could make an absolutely crackerjack game with this foundation.

Unfortunately, Mahōkenshi is just a fine game with that foundation. The balance is consistently, awkwardly off. Unlike the best Slay the Spire-alikes, synergy is not really the name of the game. A huge amount of the power of each class is just in the raw value of its best cards, so while the classes are themselves distinctive each one tends to converge on a single optimal play style rather than providing a buffet of options. And the upgrades you unlock for completing bonus challenges are so crushingly powerful that, if you get them all, the end of the game is just laughably easy.

As a design experiment, this game is fascinating. As a game to actually play, it's middling. As a bunch of French people's takes on what's most badass about Japanese mythology, it's pretty cringe. Make of that what you will.

These games are pure catnip to me. The intricate clockwork dioramas of the level designs are made tactile and enthralling by allowing the player to wreak havoc within them and see what happens. In a lesser game even this premise could wear thin after a time, but IOI knows exactly how to wring every ounce of value out of the worlds they create, from the stories woven throughout to the escalations that reframe the levels to even just the quality of writing that makes it worth listening to all the little conversations between NPCs.

The level design in 2 is generally excellent: levels exist with different organizing principles, from multiple strongholds connected by a common area like Mumbai and Miami, to layered increases in security like New York and Sgàil, to the spectacularly iconoclastic Whittleton Creek. My biggest complaint is that the DLC special assignments are lackluster, generally smaller in scope than their Hitman 1 equivalents and substantially less transformative of the levels that contain them.

I'll also mention Freelancer mode here since I played it along with Hitman 2. The underlying concept there is genius, forcing players like me who take a save-heavy approach to the game to take a wildly different approach, think on our toes, and learn to use tools that may have just sat in the loadout screen. But at the same time, failure is often more frustrating than it is fun or educational. It's just too easy to take an action that looks safe only to trigger an alarm, get gunned down, and lose your entire campaign out of nowhere.

What Freelancer needs (and the rest of the game could certainly make profitable use of) is a little bit more indication of what is or is not allowed. Show me trespassing boundaries in focus mode. Let me know whether I'm being watched before I chuck an iron at a guard's head. Guessing and checking works well enough when saves are involved, but it's not quite up to the task for a mode with such heavy consequences.

(Played Hitman 2 levels and escalations in World of Assassination)

A pretty compelling detective story. The mechanics kind of drag—there's a lot of randomly selecting and reselecting menu options just to try to get the exact set the game wants in the exact order it expects, and I can't imagine doing that without a walkthrough to get yourself unstuck periodically. But the plot itself really picks up in the latter half of the game once you start to see enough of the facts that you can start piecing together theories, and it even asks the player to make deductions without handholding a few times. A good balance of making me feel clever without making the entire thing obvious.

There's definitely a solid core of crunchy city-building within this game. The core logic of a hex-based map with various adjacency bonuses is smart, giving rise to a lot of emergent strategy out of a relatively understandable set of mechanics. But it fumbles many of the details that are necessary to grow that idea into a complete game, and ends up a frustrating shadow of what it could have been.

While my wife and I played this game together, we came up with a laundry list of complaints about it ranging from the control scheme (Why is there no hotkey for switching between planets?) to the UI decisions (Adjusting storage in a warehouse is a nightmare) to the onboarding (Why is there no Civilopedia equivalent? Why doesn't it explain core concepts like "an action"?). But ultimately, our biggest issues fell into two categories: transparency and suburbism.

The first is especially egregious in comparison to The Colonists, which is an excellent city-builder largely because it gives the player so much visibility into and control over the details of how resources flow through your civ. This is crucial for games like this, because it provides another avenue of emergent strategy, but it's totally absent in Before We Leave where you can see resources in motion, but you can't easily track them at rest or understand how their flow operates over time. It's hard to understand the paths your peeps travel even though the adjacency mechanics make this pathing critical to the player's success.

Specifically, passing near certain tiles (fountains, trees) improves the mood of a given peep while passing near others (tall buildings, pollution) makes them unhappy. Because housing imposes substantial tall building "gloom" penalties and the best food production produces pollution, you're incentivized to reproduce a suburban city layout with home, work, and food separated strictly by function. While this could pose an interesting mechanical puzzle, as a human player I want to build a city I'd like to live in. I personally love walking among mixed-use apartments, row houses, and restaurants and it's a bummer to play a game that frames that as bad.

Even the mechanical puzzle kind of sucks, though. The gloom mechanic discourages density of housing, but housing adjacency bonuses are so strong as to outweigh that completely—meaning that you just have an ambient gloom penalty all the time that you can't really address. As the game moves to later stages, this seems to be its constant theme: it keeps asking you to cram more and more things into the same number of tiles without giving you tools to improve efficiency or positive motivation to rethink layouts. (There is one notable counterexample, the step warehouse, which is worth highlighting because I wish the game had more improvements like that.)

This was a fun few hours of initial exploration, but the substantial amount of lategame content (including fighting off hostile NPCs which we haven't even talked about) just felt like a slog. And just to add insult to injury, the unlockable "reward" tiles like the Park and the Pond simply do nothing at all! What is the point.

Stylish, expressive, and funny, Hyper Gunsport is everything you could wish for in a PvP game to play with friends. Other areas of the game do feel a bit sparse—the single-player circuits are invigorating but easy to exhaust, and there's a notable absence of online matchmaking to test your skills against strangers—but if you're looking for a game to play while shooting the shit with pals, this is aces.

Although the plot is shaky at times (particularly in the third act) and the combat is rough without a light gun, this still feels like a pristine video game. The art is still gorgeous thirty years later, moody pixel art cityscapes perfectly rendering the first wholeheartedly cyberpunk video game and providing a visual legacy for the twilight of the original cyperpunk era. Even the pacing is excellent, which is quite a trick to pull off for an adventure game—a genre I dread as much as I love simply for how much time I end up spending reading item descriptions. It's a blessing that this got a contemporary release in the west.

Although I loved Death Stranding when I played it through on release, I didn't really consider it a contender for my game of the year in The Year of Endless Bangers. I let three years slip by, not even booting it back up on the release of the Director's Cut, before ultimately using my new PS5 as an excuse to transfer the save and bang out a few more deliveries.

So much has changed for me in the intervening three years. Others have written at length about the pandemic and its thematic parallels with the game, so suffice it to say that for me (still largely housebound and isolated, increasingly alienated by the fever-pitch denial of the world at large) being able to enact a world where real people work together to build infrastructure and thereby heal the world has been personally healing in a way I couldn't have imagined in 2019.

My tastes as someone who thinks critically about games have changed as well. Death Stranding's preoccupation with the texture of play—from asking you to viscerally feel the geometry of the ground you walk on to showing painstakingly mocapped cutscenes of every little action in your private room—hits much harder now that I've played through the FromSoft canon which is itself texture-obsessed in a different direction. Coming directly off playing some AAA shlock, I also found myself with a renewed appreciation of this game's dialectic approach to a cinematic aesthetic, with carefully choreographed moments that nevertheless always emphasize play as the distinguishing factor that makes this decidedly not a movie.

I'm setting this down again not because I'm finished with it for good, but because I'm inducting it into the tier of games I intend to return to over and over again. I could happily finish out the DC plots and call it "finished", but what I really want is to create a kind of personal infrastructure I can use to bring myself back to this world, this textural landscape, whenever I need to feel that connection with people that this game so masterfully evokes.

This game absolutely luxuriates in its existence as a video game. You can practically see the designers standing around a whiteboard jotting down ideas for how to use the quirks and foibles of this little adventure game engine to tell as many stories and make as many jokes as possible. The combat system is purely deterministic which might be annoying in a lesser game, but here is used to great effect both cinematically and from a design perspective—grinding is replaces by moving through the game's many subplots and exploring its areas while still capturing the joy of numbers going up.

And what better platform for a game that wants to show off its materiality than the original Game Boy? Such a tremendously underpowered little thing, barely able to hold a game together, but Kaeru no Tame manages to flip that into a strength. Every new interaction and mechanic feels like a magic trick: the game pulls a zip line out from behind your ear in the final act with a flourish, and it startles and amazes precisely because you can tell that it's at the limits of what the system can accomplish.

They don't make 'em like this anymore, and I'm not sure they even can—at least not from a major video game studio. Now it's up to hobbyists and indies to keep these platforms alive and spin their restrictions into gold.

This game's bland plot and by-the-numbers pixel art style is the least offensive thing about it. The game design is all the worst aspects of point-and-clicks, with moon logic puzzles and guess-the-pixel scenes, combined with a baffling decision not to pick up an item when you first inspect it. You have to click again, leading to many situations where you can simply fail to understand that a given object is grabbable and get stuck in a brand new mode of frustration never before seen in a genre already prone to frustration.