Idle Thoughts

A roguelike that falls short of its illustrious ancestor (★★☆☆☆)

The first Darkest Dungeon was one of the biggest indie hits of all time. It was also one of the biggest Kickstarter darlings as well. I don’t care about the exact sales number, but I’m pretty sure your game is a resounding success if multiple games copy to a t your concept of a “sidecroller party RPG with dark comic book art style” (I’m looking at you Iratus, Deep Sky Derelicts, Warsaw and Vambrace). Darkest Dungeon was also a game that did (almost) everything right: art, writing, tone, music, gameplay loop, etc.

Darkest Dungeon II is in the awkward position of having to surpass its predecessor. It fumbles right out of the gate by throwing away Darkest Dungeon’s biggest strength: it stops being a town builder/dungeon crawler to become yet another roguelike. It stumbles a second time by doing all the things a good roguelike shouldn’t do.

I wrote in a previous review on Have a Nice Death that the developers didn’t seem to have put enough thought into what makes a good roguelike tick or, even better, into why make a roguelike in the first place. Darkest Dungeon II seems to commit every cardinal sin I’ve pointed out in that review, and then some new ones.

“Roguelike” isn’t a genre, at least not anymore, now that card games, SHMUPS, sidescrollers, driving games are all roguelikes. A roguelike is a vehicle to deliver your core content, i.e. a first-person shooter, or an action game or whatever. Fine, you don’t want a regular dungeon crawler or a simple campaign mode. You want to add randomness to increase replayability and you want failure to count. Great! Now design around these ideas. We live in a post-Hades world, where that game has perfected the roguelike formula so much that the bar is now higher for everyone else.

I’ve come up with commandments. Pretentious, I know, but here goes:

– Thou shalt not make runs too long : This is the big one. Something between 20 minutes to one hour is best. The shorter the run, the less the pain of failure stings and, frankly, the less your other design flaws become apparent. In the original Darkest Dungeon (not a roguelike exactly, but still), you “won” a run every 15 minutes or so. A Darkest Dungeon II run is 2 to 5 hours long. The main reason for the difference is that in the original you pick a single dungeon. Once in a while there is also a boss. A Darkest Dungeon II run takes you through most of the areas of the game, including multiple bosses and a “final” one. This isn’t a problem in and of itself, but it makes it a lot more likely that you’ll stumble upon a tricky or just plain confusing fight. More on that later.

The other main reason for runs taking longer is the stagecoach. Basically, instead of picking nodes in a menu, you ride along in your stagecoach in full 3D and in real time. It’s slow and barely interactive. In fact, the only interactive act is riding over rubble for meager rewards, something obviously added as an afterthought to alleviate the tedium. Now, I understand why they’ve added the stagecoach. It’s atmospheric, it looks great, it exclaims “our game is in 3D now, baby!” It also increases the world’s scope: it’s not just about the Hamlet now, the entire land is in a post-cosmic horror era. More than that, it’s the entire anchor of the game: you are literally on an endless ride to redemption. To emphasize that, there is no menu on a new run, you just ride through all the “steps”: spending currency, chapter select, character selection. That setup is very annoying when you want to go back a step, though, because you can’t. Technically the ride stops when you win or lose a run, but I mean, they did try, right? But clever though the design may be, it’s still a very slow roguelike to play.

Maybe you’re thinking: in the original, losing a high-level team was a major pain and it required hours to train a new one. At least here, since this is a roguelike, win or lose, you start from scratch again, so it kind of evens out, right? Clearly, losing isn’t so bad since you’d lose that team even if you won, right? Well, the developers have a solution for you: memories! For a small cost in candles (the metaprogression currency) a character can inherit permanent bonuses between runs. Those are not enormous, but they’re not insignificant either. The rub is that to fully upgrade a character, you have to beat every chapter of the game with them, basically a 15-hour investment. Did I mention that if you either lose a run or that character dies you lose it all?  Obviously, most of the game’s bigger achievements are tied to this mechanic. Obviously. Now, I know why they came up with that idea. It’s a nice bonus for taking care of your team, a gold sink for metaprogression currency and something for the hardcore players to shoot for. But good Lord, what a way to reintroduce more frustration to the game and give the player the worst of both worlds!

– Thou shalt make your metaprogression interesting : The first Darkest Dungeon (still not a roguelike, but anyway) nailed this perfectly: between runs, you rebuild a town and recruit heroes on your crusade against evil. There were multiple resources, heroes had to be recruited independently and items you had to actually find and bring back home. I have no idea how Darkest Dungeon II messes up this bad. First of all, there is only one resource: hope, basically candles. You gather hope to rebuild an imaginary world in your mind or something? Thematically, it’s kind of confusing. Anyway, this one resource does everything: characters, items, permanent bonuses, difficulty modes, cosmetics. Very boring. The only permanent unlocks that are not tied to candles are character skills. But that’s a problem for later.

They could have thought up of something more interesting. For example, now that every character is unique (i.e., no team with multiple Plague Doctors) you could set up multiple subquests to “find” every character in the new world. Kind of like that part in Ocean’s Eleven or any heist movie where they assemble the team. A sort of “where are they now”, apocalypse edition. How about one last raid for old time’s sake, eh? You get to add world-building and secondary gameplay objectives, all in one! Now, they stumbled onto that idea with the Crusader in the paid DLC. The implementation is kind of wonky, though, because it’s somehow both vague and insistent. The game constantly pesters you with that subquest until you complete it AND it breaks one of the game mechanics (stagecoach trophies).

On the other hand, I appreciate how a few classes got major makeovers. The Bounty Hunter went back to working solo. He’ll join you semi-randomly for a region, but he’s an already unlocked, fully upgraded badass. Instead, you have to unlock his skills with candles… gotcha, more candles! The Antiquarian got demoted to being a miniboss hanging around with bandits, because she’s a greedy little…

– Thou shalt not just pile on all your unlocks in one random pool:  Most trinkets and items are unlocked with candles and just thrown in their random item pool. If you don’t like that item, well tough luck, it still makes everything else less likely to appear. Not only that, but what you’re getting is entirely opaque. You spend and you get “something”. I guess if you play long enough and spend enough candles you’ll get it all. It’s just unlocking stuff for the sake of unlocking stuff.

– Thou shalt make the game satisfying from the start : Look, you don’t need to be able to beat the game on your first run, but at least give me a solid toolset from the start. Most classes, when unlocked, lack a basic skill set for any given position. For example, the Vestal lacks one important healing skill that makes her fully effective in the back row. The Hellion lacks the skill to remove the unique debuff she inflicts upon herself that lasts all combat. This, of course, is quite annoying and makes winning a run with a new character especially hard. That’s a big design flaw when the main draw of the game is the ability to mix and match team members.

As I said, new skills are unlocked differently, through random nodes on the map. You might get 1-3 of these nodes per run, maybe. It’s not like the game even lets you pick which skill you want for a character either. Some of them are just duds. Realizing after 20 hours that I’ve barely unlocked 1-2 skills per character is probably the main reason I got tired of the game. By my calculation, it must take 40-60 hours just to see the basic abilities of each class. That is such a brazen demand of the players’ time.

The sad part is that unlocking new skills is thematically the best part of Darkest Dungeon II. Each one also gives a snippet of a character’s past. I mean, forget about the overarching story, the real heroes of this series are, well, the heroes! They all have tragic backstories and their character flaws tend to be the most obvious thing (the Hellion is a cowardly berserker, the Man-at-Arms is an incompetent officer, the Vestal is horny), but the effort is there. Each character even has a few “puzzle fights” tied to their predicament. More than that, it illustrates that only the wounded, the broken and the humbled are fit to fight the coming horrors. As the game states, the armies of the land stood there and let themselves be taken over. Only those that have already faced the darkness in themselves can face the darkness out there. Beautiful.

– Thou shalt not make the boss the only point of the game : Despite massive amounts of tooltips, it’s often not clear what’s happening during a boss fight. Even then, some fights are so unforgiving that if you allow the boss to perform its signature sucker punch move, it’s probably already too late for you. The final act bosses all have a gimmick, which may or may not ruin your hours-long run before you get to do anything about it. This, of course, directly ties into…

– Thou shalt not tempt the player into looking at the wiki : Very long runs + gimmick boss fights = frustration. The obvious way for the player to alleviate this frustration is to look up a wiki to see exactly what’s coming. You can certainly make a case for a game intended to be played wiki in hand. The game is going to kick you in the balls, but at least you have all the facts in hand, right? Even then, if so, why not make the important information part of the game itself? Why risk the player seeing more than they should and spoiling themselves? This is especially egregious in a game about mysterious, eldritch horrors. Maybe it wouldn’t matter if the game was about, I dunno, modern military hardware: you’d just be comparing the specs of main battle tanks like a good little armchair general. But you’re not supposed to just read a wiki about eldritch horrors, they’re supposed to shock you!

It doesn’t help that the first game did something similar. When I said Darkest Dungeon did “almost” everything right, one of the few things it screwed up is the endgame. The titular Darkest Dungeon at the end required some very specific knowledge and failure was punished by a very time-consuming setback. Well, this is basically the same thing here, in a slightly different form. What fool would risk hours of their time by being unprepared for a trick they know is coming?

Never mind the trick bosses, however, the game doesn’t even do that great of a job at explaining the basics. Most classes have one new gimmick that’s not obvious at first glance. For example the Vestal generates “Conviction” every turn that gets used up to empower certain skills. Those mechanics become clear eventually, but I have no idea why the game just doesn’t tell you this outright in a tutorial screen. Every character also has variants, but good luck figuring out the exact differences with in-game info. So yeah, even getting a basic grasp of the classes requires a look at the wiki.

The devs announced a new Kingdom mode, which seems like it’s bringing back the original’s Town building/Dungeon crawling structure. In other words, it is getting rid of the roguelike aspect. I’m interpreting this as an admission of guilt.

Now to be fair, the core combat loop of the game isn’t bad. It might even be better than the original. The original already took the basic RPG combat formula and improved it with tweaks like stress and positioning. Here, massive amounts of effort have been put into “anti-stalling”: i.e. to prevent prolonging fights to eke out a bit more healing. Now healing skills only work under a certain threshold and characters heal passively, albeit slowly. I’m not sure this was that big of a priority, but it is a welcome change. Some stats are gone, like protection and dodge, and most statuses like this are handled through “tokens”. So for example attacks do not miss, unless some visible status effect is in play. It doesn’t remove randomness in a game that has tons of it, but at least it makes it more readable. On the other hand, it turns fairly straightforward item effects into programming word salads. Something simple like +10% dodge chance becomes: “On hit: 12,5% chance to add 50% dodge, 12,5% chance to add 75% dodge.” Fun! Longer runs also mean you can fight multiple bosses and you are often in a position to pick your fights. To be clear, that’s a good thing. To reiterate, there is nothing wrong with the main RPG combat mechanics of the game. This brings again the question: “Why make a roguelike in the first place?”

The biggest step up compared to the original is probably the jump to full 3D. The original had no animations to speak of. Basically a camera zoom-in and the sprite tilting gave you the illusion of movement. The game still retains its top-notch dark “adult” comic book style. The game could sell itself on the strength of its visual style alone. Monsters are still stamped with a big “Deathblow” mark, making each victory very satisfying. Each hero still has a distinctive look and is now even better with animations. The Plague Doctor breaks vials that erupt in smoke. The Crusader’s parchment catches fire and drifts in the wind. Monster design is still deviously creepy. The final bosses are all body part themed, but they’re nothing as obvious as, I dunno, a Heart of Darkness. Sorry Joseph Conrad. You’ve never seen a pair of sentient lungs burst in a roar of indignant agony like this. Then again, if you asked me what to improve from the first Darkest Dungeon, “animations and 3D” wouldn’t exactly be my first pick.

The quality of the writing seems to have taken a drop as well. Wayne June has a fabulous masculine voice, the kind that could sell whisky to a teetotaller. But you still have to write great lines for a great voice. The original Darkest Dungeon is one of the most quotable games ever. I could watch the intro again and again. The original Darkest Dungeon is also one of the few “Lovecraftian” games to actually nail down the tone. In fact, perhaps the flowery prose actually out-Lovecrafts the actual author. Lovecraft wrote in a formal, wooden tone, the tone of scientists and rational men. He usually kept his fancy wording for the monsters and the stranger things out there.

Darkest Dungeon II isn’t as quotable as the first game. There’s no oxymoron as pregnant with potential meaning as “I began to tire of conventional extravagance”. There’s no sentence to underline a basic game mechanic as ridiculously prim as “Curious is the trapmaker’s art, his efficacy unwitnessed by his own eyes”. There’s no quote to keep in mind while camping like “The cost of preparedness – measured now in gold, later in blood”.

To be fair, Darkest Dungeon II is still vastly better written than most games. “The leaves fall, never to grow back again.” is a good line for its melancholic ominousness. “Spend what you can, for wealth no longer has any meaning… if, indeed, it ever did.” is a good one too.

On the other hand, the sequel makes the questionable choice of happening inside the protagonist’s head, to some unclear extent. Yeah, it’s one of those stories. Obviously, when the plot is “just a dream” or the like, the stakes are unclear. Are the heroes even real?

Other efforts were made, to varying effect. Money is no longer money. It is the end of the world after all. Money is now called “relics”… Clever, but they still function like money just fine though. The sequel is also all about your destination, a faraway mountain that gets closer with each stop. There’s a mountain in At the Mountains of Madness, right? That should be Lovecraft enough, right? Except the game has absolutely nothing to do with any mountain. It’s sort of a metaphorical mountain, get it? This fits with the overarching theme of the game of “facing your failures, despite the darkness in each of us”. It’s an interesting direction to take and the undercurrent of compassion is a nice complement to the overall bleakness of the game and the brutality of its difficulty. On the other hand, “hope will prevail” is the stuff of pop psychology, not Lovecraftian horror.

Finally, it deserves mention how bad the controls are on PS5. What I wouldn’t give to just mouse over items for help! Multiple buttons trigger tooltips (R3, L3, sometimes L2, sometimes “down”, sometimes you have to go into the pause menu). The tooltips can get in the way of making the very decision I need them for, so I have to shut them off, then get them back on the screen. Sometimes you’re stuck with a tooltip and the only way to get it to go away is to tab to another menu. The tooltips for quirks seem bugged: for some quirks, you just can’t get the game to tell you what they mean. To top it all off, another bug makes the directional buttons sometimes trigger things on completely different menus, making navigation even more difficult. It’s an awful port, UI-wise. Look, the game is basically a console RPG with a party of four. Seeing what an item does while trying to equip it is a problem that’s been solved by console RPGs for thirty years, without having to trigger tooltips multiple times. There’s no excuse. It gets easier once you start memorizing what things are, but still… I hope it gets patched. [EDIT: It got patched, partly, sort of.]

I was a Kickstarter backer for the original Darkest Dungeon. I tried to force myself to like Darkest Dungeon II out of… loyalty, I guess? They’re both very demanding games, but the sequel adds a whole bunch of questionable design decisions to the mix. You can have great fun by playing with a min-maxer mindset, wiki in hand, ready to save scum if necessary and don’t look too hard at the drop in the quality of the writing. You should also bring a lot of patience: this is a 80-100 hour game to really complete. In other words, you can make the game work for you if you have the endurance, but Darkest Dungeon II is not a “once in a generation” must-play like its predecessor. As for me, I’ll pass. We’ll see what the new Kingdom mode brings, maybe I’ll give it another shot then.