roguelikes

please-send-help-i-can’t-stop-playing-these-roguelikes.

Please send help. I can’t stop playing these roguelikes.


it’s “rogue,” not “rouge”

2025 was a very good year for my favorite genre.

Hades 2 has me in a chokehold. Credit: Supergiant Games

Hades 2 has me in a chokehold. Credit: Supergiant Games

It’s time to admit, before God and the good readers of Ars Technica, that I have a problem. I love roguelikes. Reader, I can’t get enough of them. If there’s even a whisper of a hot new roguelike on Steam, I’m there. You may call them arcane, repetitive, or maddeningly difficult; I call them heaven.

The second best part of video games is taking a puny little character and, over 100 hours, transforming that adventurer into a god of destruction. The best thing about video games is doing the same thing in under an hour. Beat a combat encounter, get an upgrade. Enter a new area, choose a new item. Put together a build and watch it sing.

If you die—immediately ending your ascent and returning you to the beginning of the game—you’ll often make a pit stop at a home base to unlock new goodies to help you on your next run. (Some people distiguish between roguelikes and “roguelites,” with the latter including permanent, between-run upgrades. For simplicity’s sake, I’ll use “roguelike” as an umbrella term).

2025 has been a truly horrific year for most things. But for roguelikes? It’s been an embarrassment of riches. Because I’m an editor and there’s no one here to stop me, I’d like to tell you about them. To keep things manageable, I’ll stick to games that hit 1.0 in 2025.

Hades II

Screenshot of hades 2

Credit: Supergiant Games

Where else could we start? In a year of wall-to-wall video game showstoppers, Hades II sticks out. The first Hades got our nod for best game of 2020, and Hades 2 certainly has my vote for 2025.

This time, you play as Melinoë, sister to Hades protagonist Zagreus and daughter of Hades himself, as she attempts to take back the house of Hades from Chronos, the titan of time. The cast of Olympian gods returns to bestow blessings (upgrades to your various attacks and defensive maneuvers) to help you on your way. If you played the first game, you’ll know what you’re getting into here; the sequel just vastly expands the content and mechanics.

As you fight through the game’s two different paths, you’ll slowly uncover the game’s story via little snippets of dialogue (there’s a truly mind-boggling amount of dialogue in this game), and oodles of unlockables and endgame challenge runs ensure you’ll be playing for a long time.

You won’t find many roguelikes with higher production values. The game is $30. Madness! If you like roguelikes, you’ve probably already picked this up. I’ll go further, though. If you enjoy video games at all, you should buy Hades II. It’s that good.

Ball x Pit

ball x pit screenshot

You ever boot up a new game and immediately think, “Well, this thing is going to be a problem for me”? Yeah.

We’ve been blessed with several pachinko-style roguelites over the past couple of years (Peglin, Ballionaire, and Nubby’s Number Factory are all worth your time); now comes a take on another ball-centric classic. I’m talking about last month’s Ball x Pit, a roguelite version of Breakout. Or at least that’s the simple way to describe it. In actuality, the game is that rarest of finds: something that feels unique.

Take one of your many and varied characters onto the battlefield, and you’ll lob a stream of balls toward the top of the screen, where slowly descending enemies periodically fire attacks back at you. When you level up, you’ll choose “special balls,” which have all manner of effects, like inflicting fire or poison on enemies or balls that explode into other balls. As the game progresses, you can “fuse” these balls together, combining the effects. Sometimes, you’ll be able to “evolve” two balls into an entirely new type of ball. Not enough for you? Slam two evolved balls together for even more wackiness.

The moment-to-moment gameplay is fantastic, with different characters and upgrades forcing you to play differently to succeed. The game doles out new mechanics and surprises along the way to keep things fresh, though this is a game you can “complete”; the between-run metaprogression eventually lets you become a bit of a god.

A base-building system—and a minigame in which you bounce your characters around the map to activate buildings—is a nice, thematic diversion between runs, but it’s mostly just a flashy upgrade screen. I usually just wanted to get back into the game as soon as possible.

Need more convincing? Check out the free demo.

Absolum

absolum screenshot

Absolum’s well-regarded demo was released in June, but this thing came out of nowhere for me. The elevator pitch: a beat ‘em up, but make it roguelite.

Not really a beat ‘em up fan? Me neither. Doesn’t matter. The last side-scrolling brawler I played for more than an hour was probably 1991’s The Simpsons arcade game or that same year’s Teenage Mutant Ninja Turtles: Turtles in Time (though the original TMNT arcade game from 1989 was my first quarter-munching arcade love).

The game features gorgeous hand-drawn visuals and the year’s best soundtrack, including this absolute banger from Doom’s Mick Gordon (that’s from a particularly metal boss fight; most of the songs are more fantasy-forward).

Over your runs, you’ll traverse multiple paths, finding secrets and unlocking new features. The roguelike elements are a bit thin at first, but piecing together powerful builds becomes simple as you fill out your options. The combat is sublime—you can get away with button-mashing for a while, but you’ll want to learn at least some of its intricacies to make it to the end.

This thing’s the complete package, and it’s easily one of the best games of the year. Bring along a friend if you’re into co-op. The surprisingly beefy demo is still available—there’s really no excuse not to check it out.

Clover Pit

clover pit screenshot

I’ve never actually been to a casino, but judging by the electric dopamine surge I get when hitting a jackpot in the slot machine roguelike Clover Pit, I know I should maintain my chastity.

Clover Pit locks you in a disgusting, blood-soaked closet of a room, and the only way to earn your freedom is to deposit money into an ATM over a series of ever-increasing payments. In other words, it’s what’s come to be known in some circles as a Balatro-like, aka a numbers-go-up game (of course, Balatro was just the first such game to hit it big; the genre’s true progenitor came a year or so earlier in the form of another excellent slot machine roguelike, Luck be a Landlord).

Standing on a trap door that will drop you to your death if you fail to hit your deadlines, you’ll pull a lever on a slot machine over and over, hoping to hit it big. It’s not totally random, though, of course. Purchaseable trinkets allow you to manipulate your odds, trigger beneficial effects, and multiply your score. Getting a jackpot of all 7’s? It’s easier than you may think.

Don’t expect Balatro-like depth—most strategies here involve simply picking one symbol and buffing it to high heaven—but fun, game-breaking builds are easy to put together to make you feel like a winner. There’s something disconcertingly hypnotic and soothing about repeatedly pulling a slot machine lever—it’s best to do it here, where you won’t end up losing your home.

Shape of Dreams

shape of dreams screenshot

I’ve been playing the hell out of this game, but if you stopped me on the street, I could not tell you what it’s called. Forgettable name aside, I love it.

I’ve heard the game’s combat and controls described as MOBA-like. That seems reasonable, at least from what I remember from my ill-advised and short-lived attempt to get into Dota 2 a decade ago. Don’t let that scare you off, though; this is basically a top-down action RPG where you’ll be fighting through small rooms of enemies, Hades-style.

What makes it special is its skill system. You start each run with a couple of attacks and a passive ability, and you’ll pick up (and replace) skills as you go. Each skill—here called “memories” (don’t ask me; I skipped the lore)—can fit up to three “essences,” modifications that affect how the skill functions. You can rearrange these modifications at any time, enabling a “make your own skill” system that’s endlessly fascinating.

Unique unlockable characters and robust metaprogression skill trees will keep you busy for embarrassing amounts of time. You can even play with friends. Before long, you’ll be creating game-breaking, room-nuking builds, the hallmark of my favorite games of the genre.

Megabonk

Megabonk screenshot

As its name might suggest, Megabonk is not a serious game. Unlike some other games on this list, there’s no chance of this one becoming an all-time great. But there’s a reason this buzzy little title has been on many gamers’ lips since its release in September. The concept is simple: it’s Vampire Survivors meets Risk of Rain 2.

If you’ve played both games, mash them up in your mind and you’ll know exactly how this thing plays. And not just in the way that most “Survivor-likes” tread extremely similar ground to the original. Megabonk‘s treasure-chest-opening animation is ripped straight from Vampire Survivors, and the game’s structure and items (down to the artwork style) are basically just Risk of Rain 2.

So no, it’s in no way original. And I was disappointed to learn that there are only two real “stages” to play; a Risk of Rain-style teleporter just takes you to a harder version of the stage you’ve picked. There are also balance issues; the damage scaling on anything but the first 10-minute stage is absurd. But there is some ridiculous fun to be had with it.

If you’re not into the whole auto-shooter/bullet heaven thing, there’s nothing here for you. But if you’re interested in seeing how chaotic a third-person Vampire Survivors can get, step right up.

It’s also the one 2025 game where you can play as a sunglasses-wearing, skateboarding skeleton who throws bouncing bones at enemies. In these tough times, that’s not nothing.

Deep Rock Galactic Survivor

deep rock galactic survivor screenshot

We’ve talked about this Survivors-like take on the beloved co-op shooter Deep Rock Galactic a couple of times over its Early Access period, but we were remiss in not discussing it upon its 1.0 release last month. The game was already an Ars favorite, but its progression systems still needed a bit of work. It’s now ready for public consumption, and it’s one of the best auto-shooters on the market. It’s so good that you might want to take a look at it even if you want nothing to do with the oversaturated subgenre.

Its Vampire Survivor-like bones are obvious—you walk around a map while your weapons fire automatically at hordes of enemies closing in on you. Collect the XP gems defeated enemies drop to level up and choose an upgrade. The difference here is that you’re also able to mine through walls of rocks, letting you escape tricky situations and funnel bad guys to traps you’ve laid.

The progression system is heavy on the grind, but there’s plenty of fun to be had no matter how hardcore you want to be about it.

Rock and stone!

Monster Train 2

monster train 2 screenshot

Five years after the original, it’s time for the sequel to the second-best roguelike deckbuilder of all time (the sequel to the first-best roguelike deckbuilder has—thankfully, if I’m being honest—been delayed until the beginning of next year). As in the first game, and as the game’s title might suggest, you’ll be fighting monsters on a train, trying to stop them before they ascend three floors to reach your “pyre”—your health pool for the run.

In Monster Train 2, as in any deckbuilder, you start with a fairly crappy deck of cards and upgrade and expand it throughout your run to try to make it to the end. But in addition to the usual spells and attacks, Monster Train 2 gives you units to assign to the different levels of your battlefield, infusing an interesting spatial element to the cartoonishly violent proceedings.

The sequel is more of the first game, but with smart updates that make everything flow smoother. It’s one of my favorite games of the year, and I highly recommend it to any fan of tactical card games.

Deadzone Rogue

deadzone rogue screenshot

Deadzone Rogue instantly joins the pantheon of roguelite first-person looter shooters, which includes perennial favorites Gunfire Reborn and Roboquest (I haven’t played them yet, but the brand-new Abyssus and Void/Breaker are also generating a bunch of buzz).

Where Roboquest excels at fun, Doom-like movement and colorful environments, Deadzone Rogue is all about the shooting. The game has the best gunplay of any FPS roguelike I’ve played, and the random weapons, armor, and upgrades you get give each run a sense of personality.

The game’s music, voice acting, and lore are best ignored, but the sound design is nice and punchy. This won’t be a game you’ll play for 100 hours, but sometimes it’s just fun to shoot a gun in a video game, and Deadzone Rogue gets that simple formula right.

9 Kings

9 kings screenshot

Look, I’m going to cheat here, and I’m not ashamed of it. It’s true—9 Kings is not fully released. But I can’t not talk about. I initially wrote the game off when it was released into Early Access in July, thinking it looked too simple. It is simple, but that’s to its credit.

The premise is easy to explain: Build a little kingdom on a 3-by-3 grid of squares. Play a card to construct or upgrade a building or unit in your kingdom. Afterward, a neighboring kingdom will attack, and your units will automatically fight to defend your home. After the battle, you draft a card from the defeated kingdom to add to your hand.

As you can see from the above screenshot, you can expand your kingdom beyond the initial nine squares, and unlockable perks change up the way you play each king. A handful of enemies are randomly chosen from the pool of nine, meaning that the cards you can draft each run will be different.

Making busted builds and fighting your way up the difficulty levels is extremely compelling; there was a week where the “one more run” curse descended on me, and I did little else than play this game.

Photo of Aaron Zimmerman

Aaron is Ars Technica’s Copy Chief. He has worked as an editor for over 17 years. In addition to editing features at Ars, he occasionally reviews board and video games. He lives in Chicago.

Please send help. I can’t stop playing these roguelikes. Read More »

what-kind-of-bug-would-make-machine-learning-suddenly-40%-worse-at-nethack?

What kind of bug would make machine learning suddenly 40% worse at NetHack?

Large Moon Models (LMMs) —

One day, a roguelike-playing system just kept biffing it, for celestial reasons.

Moon rendered in ASCII text, with

Aurich Lawson

Members of the Legendary Computer Bugs Tribunal, honored guests, if I may have your attention? I would, humbly, submit a new contender for your esteemed judgment. You may or may not find it novel, you may even deign to call it a “bug,” but I assure you, you will find it entertaining.

Consider NetHack. It is one of the all-time roguelike games, and I mean that in the more strict sense of that term. The content is procedurally generated, deaths are permanent, and the only thing you keep from game to game is your skill and knowledge. I do understand that the only thing two roguelike fans can agree on is how wrong the third roguelike fan is in their definition of roguelike, but, please, let us move on.

NetHack is great for machine learning…

Being a difficult game full of consequential choices and random challenges, as well as a “single-agent” game that can be generated and played at lightning speed on modern computers, NetHack is great for those working in machine learning—or imitation learning, actually, as detailed in Jens Tuyls’ paper on how compute scaling affects single-agent game learning. Using Tuyls’ model of expert NetHack behavior, Bartłomiej Cupiał and Maciej Wołczyk trained a neural network to play and improve itself using reinforcement learning.

By mid-May of this year, the two had their model consistently scoring 5,000 points by their own metrics. Then, on one run, the model suddenly got worse, on the order of 40 percent. It scored 3,000 points. Machine learning generally, gradually, goes in one direction with these types of problems. It didn’t make sense.

Cupiał and Wołczyk tried quite a few things: reverting their code, restoring their entire software stack from a Singularity backup, and rolling back their CUDA libraries. The result? 3,000 points. They rebuild everything from scratch, and it’s still 3,000 points.

<em>NetHack</em>, played by a regular human.” height=”506″ src=”https://cdn.arstechnica.net/wp-content/uploads/2024/06/13863751533_64654db44e_o.png” width=”821″></img><figcaption>
<p><em>NetHack</em>, played by a regular human.</p>
</figcaption></figure>
<h2>… except on certain nights</h2>
<p>As <a href=detailed in Cupiał’s X (formerly Twitter) thread, this was several hours of confused trial and error by him and Wołczyk. “I am starting to feel like a madman. I can’t even watch a TV show constantly thinking about the bug,” Cupiał wrote. In desperation, he asks model author Tuyls if he knows what could be wrong. He wakes up in Kraków to an answer:

“Oh yes, it’s probably a full moon today.”

In NetHack, the game in which the DevTeam has thought of everything, if the game detects from your system clock that it should be a full moon, it will generate a message: “You are lucky! Full moon tonight.” A full moon imparts a few player benefits: a single point added to Luck, and werecreatures mostly kept to their animal forms.

It’s an easier game, all things considered, so why would the learning agent’s score be lower? It simply doesn’t have data about full moon variables in its training data, so a branching series of decisions likely leads to lesser outcomes, or just confusion. It was indeed a full moon in Kraków when the 3,000-ish scores started showing up. What a terrible night to have a learning model.

Of course, “score” is not a real metric for success in NetHack, as Cupiał himself noted. Ask a model to get the best score, and it will farm the heck out of low-level monsters because it never gets bored. “Finding items required for [ascension] or even [just] doing a quest is too much for pure RL agent,” Cupiał wrote. Another neural network, AutoAscend, does a better job of progressing through the game, but “even it can only solve sokoban and reach mines end,” Cupiał notes.

Is it a bug?

I submit to you that, although NetHack responded to the full moon in its intended way, this quirky, very hard-to-fathom stop on a machine-learning journey was indeed a bug and a worthy one in the pantheon. It’s not a Harvard moth, nor a 500-mile email, but what is?

Because the team used Singularity to back up and restore their stack, they inadvertently carried forward the machine time and resulting bug each time they tried to solve it. The machine’s resulting behavior was so bizarre, and seemingly based on unseen forces, that it drove a coder into fits. And the story has a beginning, a climactic middle, and a denouement that teaches us something, however obscure.

The NetHack Lunar Learning Bug is, I submit, quite worth memorializing. Thank you for your time.

What kind of bug would make machine learning suddenly 40% worse at NetHack? Read More »

why-there-are-861-roguelike-deckbuilders-on-steam-all-of-a-sudden

Why there are 861 roguelike deckbuilders on Steam all of a sudden

A very full house —

9 answers from 8 devs about why combat card games on screens have blown up.

A hand holding a set of cards from popular roguelike deckbuilders, including Slay the Spire and Balatro

Aurich Lawson

In a deckbuilding game, you start out with a basic set of cards, then upgrade it over time, seeking synergies and compounding effects. Roguelikes are games where death happens quite often, but each randomized “run” unlocks options for the future. In both genres, and when they’re fused together, the key is staying lean, trimming your deck and refining your strategy so that every card and upgrade works toward unstoppable momentum.

“Lean” does not describe the current scene for roguelike deckbuilder games, but they certainly have momentum. As of this writing, Steam has 2,599 titles tagged by users with “deckbuilding” and 861 with “roguelike deckbuilder” in all languages, more than enough to feed a recent Deckbuilders Fest. The glut has left some friends and co-workers grousing that every indie game these days seems to be either a cozy farming sim or a roguelike deckbuilder.

I, an absolute sucker for deckbuilders for nearly five years, wanted to know why this was happening.

  • In Slay the Spire, and most roguelike deckbuilders, you battle enemies by drawing cards and playing a limited number of them: attack, defend, buff, debuff, etc. Crucially, enemies show you what they are going to do next, so you can triage and strategize.

    MegaCrit

  • Winning battles nets you random new card choices, which may or may not fit your strategy.

    MegaCrit

  • You choose which path to take, full of battles, stores, random “encounters,” rest stops, and “Elite” battles that are more rewarding.

    MegaCrit

  • Stores and encounters will often let you buy cards or artifacts, and sometimes remove them, too.

    MegaCrit

  • Roguelike deckbuilder bosses are often designed to challenge build strategies and force adaptability.

    MegaCrit

  • Or, as often happens, you just die and start over with more cards and upgrades unlocked for next time.

    MegaCrit

What is so appealing to developers and players about single-player card games made for screens? How do developers differentiate their deckbuilders? And how do you promote a title in a niche but crowded field?

Seeking these answers, I spoke with a bunch of roguelike deckbuilder developers, and I read interviews and watched conference talks from others. Some common themes and trends revealed themselves. Like a well-honed deck, each element fed into and bolstered the others.

But let’s first go back to the beginning, to perhaps the most powerful single element of roguelike deckbuilders’ success: two college friends in their 20s, tired of working QA jobs.

Slay the Spire’s starting point

Slay the Spire marked what was arguably the start of modern, single-player roguelike deckbuilder video games. Some games may technically have combined combat-oriented deckbuilding with the procedural generation and die/improve/repeat nature of roguelikes, but the 2019 game was the first to crack the formula and build a big audience around it. Slay the Spire also broadly boosted enthusiasm for single-player card games on computers in general—games other than Windows’ Solitaire, at least.

Video directed by Justin Wolfson, edited by John Cappello. Click here for transcript.

In a video interview with Ars Technica, and at Game Developer Conference (GDC) talks in 2019 on marketing and balancing, developers Anthony Giovannetti and Casey Yano told the game’s story. Giovannetti and Yano had met in college and made some one-off games, then graduated and got jobs. Giovannetti was a card game and tabletop enthusiast, even briefly managing a game store. He was certainly familiar with deckbuilding pioneer Dominion, but his main game was Netrunner—he still maintains the community site StimHack. Yano worked at Amazon, where he said he picked up the company’s “customer obsession” mentality.

In mid-2015, the two reconnected and went all-in on making their genre-melding concept, initially named “Card Crawl.” Starting with stick-figure drawings, a procedurally generated progression scheme cribbed from FTL, and input from some advanced Netrunner playtesters, they worked until the game was ready for early access on Steam. Chief among their in-development discovery was broadcasting enemy intents to the player and simplifying visuals and indicators until they were readable at a glance, even in a foreign language.

Slay the Spire launched in Steam’s Early Access after more than two years of development in November 2017. It sold 200 copies on day one, 300 on day two, and 150 on day three, declining from then on. The developers had made trailers, sent more than 600 emails to press and other outlets, and in the critical first two weeks of release, they had only sold 2,000 copies.

Things looked grim, but eventually, some of the 200 keys they sent to streamers led to some live play. An influential Chinese streamer’s Slay session garnered more than 1 million views, which nudged the game up the top seller list, leading to further sales, which sparked more streams, and so on. Grateful for their second wind, the team released new patches every week and used statistical feedback from early sessions to further tune the game. They took care not to remove “overpowered” strategy discoveries because they understood the joy of “a well-powered Rube Goldberg machine.”

Despite critical raves, a 99 percent positive Steam review rating, and more than 1.5 million sales by September 2019, Yano told the GDC crowd that “we never really improved how to, like, sell the game. I would say it’s still really word-of-mouth. But it’s been doing well that way, so I think we’re gonna keep going that way.”

Multiple developers I spoke with cited Slay the Spire as inspiration; one had more than 1,000 hours in it. The game’s design and success have compounded a few times over, creating new starting points. Balatro‘s developer claimed to have not played deckbuilders before making his own, but he was fascinated by streams of Luck Be a Landlord. That slot machine roguelike was, per its developer’s blog, heavily influenced by Slay the Spire. Even if you don’t know it, you probably know it.

<em>SpellRogue</em>, from a two-person team, has cards, but you use them by rolling dice and fitting the results into the cards’ slots (Yahtzee!).” height=”1440″ src=”https://cdn.arstechnica.net/wp-content/uploads/2024/04/Screenshot_EarlyAccess_3840x2160-2-scaled.jpg” width=”2560″></img><figcaption>
<p><em>SpellRogue</em>, from a two-person team, has cards, but you use them by rolling dice and fitting the results into the cards’ slots (Yahtzee!).</p>
<p>Guidelight Games/Ghost Ship Publishing</p>
</figcaption></figure>
</p></div>
<nav>Page: <span>1 <a href=2 3 4 Next

Why there are 861 roguelike deckbuilders on Steam all of a sudden Read More »

cobalt-core-is-a-tight,-funny-roguelike-deck-builder-deserving-of-many-runs

Cobalt Core is a tight, funny roguelike deck-builder deserving of many runs

Gravitational pull —

It’s got style, dialogue, and even a plot on top of procedural card battling.

Scene from Cobalt Core game

Enlarge / These symbols might not mean anything to you now. But give it a few runs, and you might lose a few minutes strategizing this ship’s ideal next turn, based on this image alone.

Brace Yourself Games

Games come and go through my Steam and Nintendo Switch libraries: a twitchy, grim action epic, then a metaphysical puzzle-platformer, and maybe a boomer shooter or turn-based tactical along the way. I try hard not to get stuck in one style or mindset—both for my enjoyment and my writing.

But there is always one type of game that is installed and ready to go for the next trip or idle couch moment: a roguelite deck-builder. Cobalt Core is the latest game in that slot, and it’s on Steam for Windows (and definitely Steam Deck) and Switch. It’s the most fun I’ve had in this particular obsession since Monster TrainCobalt Core stretches into other genres, like perfect-knowledge turn-based tactics and space battle, but it’s cards and randomness down to its electric-blue center.

Launch trailer for Cobalt Core.

A few years ago, I didn’t know what a “roguelike deck-builder” was or what either of those compound phrases meant. Then, one day, there was a sale on Slay the Spire. That 2019 game refined the fusion of two game mechanics: constant failure against randomized encounters (a la Rogue, but with a “lite” gradual progression) and the refining of a deck of combat-minded cards (as in Magic: The Gathering, Dominion, and Netrunner). You attack and defend against increasingly tough enemies with your cards, you gain and upgrade and ditch cards as you go, you lose, and then you get slightly better tools on your next do-over.

Done well, roguelike deck-builders are a potent mix of luck, immediate and long-term strategy, and the slow dopamine drip of chained-together victories. I’ve lost entire work months to them, cumulatively. They should come with a warning label.

Cobalt Core has all those addictive elements plus a few more things to do with your cards. For one, there is space jockey positioning. You move your ship left and right against your opponents, lining up beam emitters and missiles, targeting weak points, and dodging. You do this with perfect Into the Breach-like knowledge of the opposing ship’s actions: it’s going to fire from this bay for 2 damage, use this bit to upgrade shields, and so on. This leaves you with the decisions of when to take hits versus dodging, where to launch drone cannons, whether to attack now or wait for better cards next hand.

  • A mid-battle scene, with your ship (on bottom) in the midst of delivering a hull-smashing attack, just before the enemy’s multiple red beams rain down.

    Brace Yourself Games

  • You pick your battles in Cobalt Core, and also choose repair shops, random happenings, and mid-level bosses.

    Brace Yourself Games

  • The cards you’re offered depend on which cremates you picked for this run, and each has a distinct discipline.

    Brace Yourself Games

  • The cover art for Cobalt Core, which makes you really not want to get your new friends blown up.

    Brace Yourself Games

On top of working those elements into fine shape, Cobalt Core cleverly embraces them in its plot and theming. Your ship is made up of quirky characters, each of whom adds their cards with unique play styles to your deck. You fight ship after ship, encounter celestial oddities, warp from sector to sector, then fight the titular object. When you beat it, you … do something, in space and time, it seems, then start over with imperfect memories of prior loops. Each win or notable loss unlocks new crew, cards, and memories, enough of which might explain what exactly is going on in this heady quantum plot.

What’s largely drawn me in is how neatly the game’s battles and runs fit into a casual gaming schedule. A single battle is usually less than five minutes, a sector of fights maybe 15–20, and a couple sectors plus a boss fight 30–40 minutes, though the decisions and early sections get faster as you learn them. You can save and exit anytime, even mid-battle, and it’s a fast-loading game. On the Switch or Steam Deck—for which this game is Verified and a real battery-saver—it makes for a generous couch, waiting room, or travel experience.

I’ve still got a lot to unlock in Cobalt Core after 10 hours of play, and could easily see myself hitting the 100-plus I put into Slay the Spire and Monster Train (or at least the 50 I threw into Inkbound or SteamWorld Quest). It helps that there’s much more singular personality and style in Cobalt Core than other procedural card battlers. And humor, too, lots of it, accompanied by appealing 16-bit-esque graphics, making the grind for new cards and esoteric achievements feel far less grinding.

It’s also impressive how much character-building the game pulls off in service of pixel portraits that never move but frequently react, quip, and express their quantum confusion. I typically feel nothing for different characters in these games. But I’ve got a soft spot for the gunner Riggs, and feel bad when I skip him for, say, movement or hacking options.

My major complaint about Cobalt Core, this deep into its clutches, is that its soundtrack is catchy. This is normally a positive, but given how regularly I’m dropping in for a round or two or 10, I find my mental synth deck replaying the melody lines from a few different tracks. I can’t hum a single note from any of the orchestral-minded backings of my prior deck-builder obsessions, but these tunes are burrowing deep inside.

Cobalt Core is currently on sale (on both platforms), and its base price is $20–$25. If you know you like picking cards, beating bosses, and a long, meticulous triumph, I have to imagine it’s a great value. Maybe too much of one.

Cobalt Core is a tight, funny roguelike deck-builder deserving of many runs Read More »