Slay the Spire 2 review

written 2026-09-12

Ironically, though Slay the Spire pioneered the roguelike deckbuilder genre, I only played it after two other roguelike deckbuilders (Hellcard and Cobalt Core), and only its sequel. I got into it in 2026 and first won its hardest setting legitimately on 2026-09-11.

Combat

Combat consists of playing cards to generate damage and block, and enemies indicate their actions a turn ahead (so if they're going to attack for X, you want at least X block this turn).

Sadly, combat often plays itself: play your block cards from best to worst until you have enough block, then play your attacks from best to worst. This is especially true in the first few fights where most of your deck is vanilla cards that just give block or attack with no other effects or interactions.

Related: this game starts you with more of these cards than other deckbuilders I've played, and gives fewer opportunities to remove them, so you'll be drawing them for the whole run.

There's also much luck, because draw order matters much: block doesn't persist across turns, so if you draw block cards on turns the enemy isn't attacking, they're useless, and then you don't draw them when the enemy is attacking, so you take a bunch of unavoidable damage. Your deck is also bigger than other roguelike deckbuilders I've played, so having permanent buffs shuffled to the bottom hurts much more. This is mitigated by your high max HP so a little bad luck won't kill you, but it does hurt.

There are potions, consumables items that can be used in combat, which mitigate both these problems, but they're not a big enough part of the game to make up for lacking core mechanics.

Building

Building mostly consists of choosing one of three cards to add to your deck after each fight. And this aspect is good, because card balance is good; there are few over- or underpowered cards, so this decision is usually interesting.

There are relics, passive bonuses that don't need to be drawn, but they mostly come from treasure chests and elites which only give you a random one, no choice, so they don't add strategic depth. Instead, they add luck, since you might get one very good or one very niche and useless for your build, and have no control over it.

You also get small amounts of gold from each fight, and can spend it at a shop for cards, relics, potions, or removing a card. The shop is a good mechanic because it's the only place that lets all kinds of items compete for your choice, but you usually only get 3-4 per run. And buying cards is usually bad because you get so many card rewards anyway. And the relics are often too expensive to afford.

Finally there are random events which can offer all sorts of choices, mixing outcomes like adding, upgrading or removing cards, adding curses (negative cards), gaining or trading relics, and gaining or losing HP or max HP. I like how they often have both an upside and a downside. But again I dislike how much luck these introduce, not only because it randomly chooses from an incredibly wide set of options, but because the random event nodes on the map aren't always random events, sometimes they're just normal fights, sometimes they're shops, sometimes they're treasure chests (extremely good).

Card design

My only complaint about card design is that there's many whose value can't be evaluated well because they randomly choose from too wide a set of outcomes. Many cards and potions do things like "add a random attack to your hand". The value of this is the average value of all attacks in the game, which just can't be evaluated except by massive experience.

Routing

There's a journey map where you choose a path between nodes of different types like "normal enemy", "elite enemy", "random event", "rest site", and "shop". This is very similar to Cobalt Core and at least better balanced than that, but still not great. You always want as many shops as you can afford and as many rest sites as possible, since these offer reliable ways to improve your deck (at rest sites you can upgrade a card if you don't need healing). Normal enemies only give a tiny amount of money and a random card reward, which you'll often refuse. And usually you'll take at least some damage in each fight. And random events are almost always better than fights, but worse than shops and rest sites.

Routing would be much more interesting if you could see which specific enemy is in each enemy node so you could strategize by avoiding enemies you're weak against or building around enemies you won't be able to avoid. But you can't, so it's just luck whether you get an enemy you're strong against or not.

You can see what the boss of each act is going to be, but only once you're in that act. Different act 3 bosses have very different counters, so you need to build for the boss you get, but by the time you see what it'll be, there's only one act left and not enough opportunity to change your build.

Learnability

Compared to other roguelike deckbuilders I've played, there's much more content: more cards, more relics, more enemies, more events. On one hand this is great because it adds replay value for experienced players, but on the other, it lengthens the unfair learning period. For example, one of my first few runs had a final boss called "Aeonglass" which I'd never seen before, I had an very powerful deck which was crushing everything else but it turned out Aeonglass had a unique mechanic that hard countered it. The same happens with elites and to some extent with normal enemies - experts always say you need to know the pool of possible fights in each act and plan around how well you could handle each one, and also know the exact move cycle of each enemy to plan turns around, but that's not possible until you have a hundred hours of experience, or if you read the wiki during runs.

Configuration

There's not much run configuration. The only parameter is which character to play as, which there's 5 of. And the characters are cool - each one has several unique mechanics that add much to the game. But it's not as interesting as the configuration options in the other deckbuilding roguelikes I've played.

There is a "custom" run mode with several other settings, but it seems like more of a silly novelty than a serious game mode, because many of the settings do absurd things that drastically change and imbalance the game, like completely replace your starting deck with 10 card rewards, or triple the amount of gold you get but disable upgrading cards at rest sites. To be clear, I think it's cool that this mode exists, but it doesn't provide the appeal of the configuration options in Hellcard, Cobalt Core, or Invisible, Inc..

Locking ascensions

There are 10 levels of hard mode, called ascensions, which all have to be unlocked. Locking hard modes is a toxic practice in general, but here is the worst case I've seen, because you have to unlock them *separately for each character*, and then separately in multiplayer. When the differences between each level are so small and the base level isn't hard for someone experienced with the genre, this is incredibly tedious. I had to edit my save file to unlock the ascensions, and even that took hours to figure out how and only partly worked.

Coop

This game has online coop with friends, but sadly, isn't designed around it. There are few ways to interact with your allies: you can't play defensive or power cards on them, and no fights have mechanics particularly meaningful in coop; they all just target all players with every attack. Still, this mode being available adds some value to the game.

Conclusion

Overall, Slay the Spire 2 feels like a genre pioneer in a bad way: an unpolished early form of the idea. Slay the Spire 1 should feel this way, but not 2. Its combat is shallow and it has tons of luck, but it still has a deep strategic landscape, and its high level of difficulty gives it long life.