Posted May 08, 2017
I've been playing quite a lot of Darkest Dungeon lately. In fact, I've almost completed it for a second time, and it's not an especially short game. This is unusual for me; it's a rare game that I manage to complete even once.
It's a bit of an odd game. I've heard it described as a “rogue-like”, and while it does check a few “Rogue” boxes (it's a dungeon crawl; it has perma-death, sort of, anyway; and the dungeon layouts are, mostly, procedurally generated), this isn't a very good description in general- in most respects it has little to nothing in common with traditional rogue-likes.
It also has a reputation for being hard. I think this is actually unfair, and the perception probably stems from three things that are true of it:
Firstly, it has a relatively steep initial learning curve. Going into it “blind”, without any outside advice (which I have to admit I did not do), would probably result in a lot of initial failure until some of the basics had been understood. Fortunately, at this stage the game is at its least punishing; in a worst case one loses only perhaps 20-30 minutes for really messing up completely.
Secondly, it does have a few difficulty “spikes”. Most particularly, the jump to Champion level dungeons is quite steep.
Thirdly, and most damningly, the game is absurdly punishing in its end-game. It's not “hard”, per-se; it doesn't take a lot of experience at the end-game to become good enough to essentially never fail. But it is horrendously punishing.
Fair warning from here-on: gameplay spoilers ahead. I shan't spoil the plot. The game has practically no plot to speak of anyway; it works almost entirely on atmosphere and scene-setting, something it does extremely well, I might add. But I will talk about strategy and gameplay mechanics in a way that might be considered spoiling; a good part of the game's fun is figuring things out.
I'm going to start with a criticism of the end-game. I think that, while the atmosphere is as good here as anywhere, if not better, the end-game has some unfortunate gameplay elements. It combines a few sudden changes in mechanics and a complete replacement of all enemies with new ones (both meaning that a player experienced with the rest of the game is going to suddenly encounter a lot of new things), with a new rule that prohibits retreating without suffering serious loss. And in Darkest Dungeon, losing a high-level character is very punishing, much more so than losing a low-level one. It takes roughly 8 dungeons to level a character to resolve 6. Since high-level characters can't help out in low level dungeons, and taking a very low level character into a high-level dungeon is very risky, in an extreme case, this means losing around 4 hours gameplay from a single mistake.
Unless I missed it somewhere, I don't think the game warns you that you are going to need 16 characters at minimum, either. This isn't a huge problem, since you don't lose very much by leveling, say, 12 characters and then discovering that you need another team that you now need to level up, but it is certainly easier to level all four teams at once, because you can switch between them as they get stressed.
As a sort of irony, the end-game is probably one of the easier parts of the game once you've played through it a few times. Because the end-game (and only the end-game) is entirely static; the dungeon layouts and even the enemies and their positions are all fixed, there are no surprises. This somewhat lessens replayability.
Finally, it is more than a little weird that the end-game levels actually get easier as you progress through them. I think you could argue that 2 is harder than 1 (I didn't find it so, but that may have been largely luck), but 3 is certainly easier than either (at least with spoilers) and 4 is the easiest of the lot.
I have to confess that, doing a little research and discovering how punishing the game is at this point, I did not play through the end-game blind and consequently got through it without losing any characters. But I feel in some ways my spoilered-approach was less than ideal, and especially in the third dungeon, I lost a great deal from the intended experience by not playing blind. Of course, you can only play it blind once at most anyway.
I don't have a simple solution to all of these problems. I think the end-game levels in general, and especially 3, should have been procedural, as the rest of the game is. I think the “lose a hero if you retreat” mechanic was misguided. But neither of these changes would affect the issue whereby losing a high-level hero is much more punishing than losing a low-level one.
I've read some other criticism to the effect that the game is too long. I would tend to agree, tentatively. I've completed the game once in Radiant mode (which is intentionally a bit shorter), which was just about OK. A few months later I started a new game in Stygian mode. I've reached week 32; I have 16 level 5-6 characters, but I've realised I need around 160,000 gold to upgrade all of their weapons and armour to level 5 (all guild and blacksmith upgrades are done). This is probably at least 8 weeks more before starting the endgame. I've given up for the moment; it was just feeling too grindy. Strangely, my Radiant playthrough actually took longer, at least in terms of weeks; I finished in week 56. But then, I had a few deaths, and wasn't playing quite as optimally in general.
Beyond this point, I'm going to continue to criticise (it's much more fun to point out the flaws in something than to sing its praises). But I should perhaps re-stress that I do think Darkest Dungeon is a good game; I wouldn't have played through it twice if I hadn't.
There are quite a number of what I would call “balance” problems. Curiously, this doesn't include balance between adventurer classes. I'm sure there are players with enough experience to be able to say that some classes are over or under-powered, but I don't consider myself one of them. While I've certainly formed some opinions, I can't make any confident statements about the power of any class. The dialogue I've seen in forums about this never really seems to reach a consensus from what I can see, though I appreciate that patches have changed the situation from time to time. I'm much more confident about some of my strategic assessments.
Where to start? Well, the eight buildings you can upgrade are not very well balanced. You need some Stagecoach roster upgrades, and fairly soon, too. One or two network upgrades are probably useful, too, though you certainly don't want them all. You pretty much need to upgrade the Guild and Blacksmith fully, and as soon as possible. You want a few Sanatorium upgrades, but only in the late game when diseases start to be common and only for disease treatment. One or two Survivalist upgrades might be useful (I've never gone for them, though). The Abbey and Tavern upgrades are entirely useless. The Nomad Wagon I believe is too expensive to warrant using. As I note above, money has been the biggest constraining factor for me in how quickly I can finish the game; XP and heirlooms come quicker than money.
On a related note, adventurer traits appear at first glance to be fundamental to the gameplay. But I've actually concluded that for the most part they really don't matter that much. Very few negative traits are a real problem at all (most of the “curiosity”-type traits will almost never impact negatively on you, for example. Kleptomaniac is the worst of this group, but it's really not a big problem). Even the best positive traits are fairly marginal; certainly worse than a single armour or weapon upgrade.
And there is really no good strategy available for even trying to manage traits. Removing locked negative traits is far too expensive to consider. Removing unlocked ones is still quite expensive, and the hero will likely acquire another one shortly anyway, so why bother? Locking in positive traits is also too expensive to justify. Dismissing heroes who acquire bad negative traits is far too costly in terms of XP lost. So mostly you just ignore them.
There's a curious lack of balance in terms of time spent at each of the three dungeon levels. In both my runs, nearly 70% of my runs were in Veteran dungeons. You gravitate quite quickly from Apprentice to Veteran because the increase in XP is so large, Veteran is really not that hard, and the XP gap from level 0 to 1 is so small. You're required to do Veteran dungeons from levels 3 to 5/6 anyway. But there's really very little reason to do Champion dungeons at all. The difficulty step up is huge. By the time you can possibly consider attempting them (you really don't want to be risking level 4 heroes in Champion dungeons unless you're a much better player than me) you're so close to the endgame that there's very little need to get any more XP on the hero. Resolve level 6, from what I can see offers +10% to all resists and absolutely nothing else, so it's not at all important.
It's a bit of an odd game. I've heard it described as a “rogue-like”, and while it does check a few “Rogue” boxes (it's a dungeon crawl; it has perma-death, sort of, anyway; and the dungeon layouts are, mostly, procedurally generated), this isn't a very good description in general- in most respects it has little to nothing in common with traditional rogue-likes.
It also has a reputation for being hard. I think this is actually unfair, and the perception probably stems from three things that are true of it:
Firstly, it has a relatively steep initial learning curve. Going into it “blind”, without any outside advice (which I have to admit I did not do), would probably result in a lot of initial failure until some of the basics had been understood. Fortunately, at this stage the game is at its least punishing; in a worst case one loses only perhaps 20-30 minutes for really messing up completely.
Secondly, it does have a few difficulty “spikes”. Most particularly, the jump to Champion level dungeons is quite steep.
Thirdly, and most damningly, the game is absurdly punishing in its end-game. It's not “hard”, per-se; it doesn't take a lot of experience at the end-game to become good enough to essentially never fail. But it is horrendously punishing.
Fair warning from here-on: gameplay spoilers ahead. I shan't spoil the plot. The game has practically no plot to speak of anyway; it works almost entirely on atmosphere and scene-setting, something it does extremely well, I might add. But I will talk about strategy and gameplay mechanics in a way that might be considered spoiling; a good part of the game's fun is figuring things out.
I'm going to start with a criticism of the end-game. I think that, while the atmosphere is as good here as anywhere, if not better, the end-game has some unfortunate gameplay elements. It combines a few sudden changes in mechanics and a complete replacement of all enemies with new ones (both meaning that a player experienced with the rest of the game is going to suddenly encounter a lot of new things), with a new rule that prohibits retreating without suffering serious loss. And in Darkest Dungeon, losing a high-level character is very punishing, much more so than losing a low-level one. It takes roughly 8 dungeons to level a character to resolve 6. Since high-level characters can't help out in low level dungeons, and taking a very low level character into a high-level dungeon is very risky, in an extreme case, this means losing around 4 hours gameplay from a single mistake.
Unless I missed it somewhere, I don't think the game warns you that you are going to need 16 characters at minimum, either. This isn't a huge problem, since you don't lose very much by leveling, say, 12 characters and then discovering that you need another team that you now need to level up, but it is certainly easier to level all four teams at once, because you can switch between them as they get stressed.
As a sort of irony, the end-game is probably one of the easier parts of the game once you've played through it a few times. Because the end-game (and only the end-game) is entirely static; the dungeon layouts and even the enemies and their positions are all fixed, there are no surprises. This somewhat lessens replayability.
Finally, it is more than a little weird that the end-game levels actually get easier as you progress through them. I think you could argue that 2 is harder than 1 (I didn't find it so, but that may have been largely luck), but 3 is certainly easier than either (at least with spoilers) and 4 is the easiest of the lot.
I have to confess that, doing a little research and discovering how punishing the game is at this point, I did not play through the end-game blind and consequently got through it without losing any characters. But I feel in some ways my spoilered-approach was less than ideal, and especially in the third dungeon, I lost a great deal from the intended experience by not playing blind. Of course, you can only play it blind once at most anyway.
I don't have a simple solution to all of these problems. I think the end-game levels in general, and especially 3, should have been procedural, as the rest of the game is. I think the “lose a hero if you retreat” mechanic was misguided. But neither of these changes would affect the issue whereby losing a high-level hero is much more punishing than losing a low-level one.
I've read some other criticism to the effect that the game is too long. I would tend to agree, tentatively. I've completed the game once in Radiant mode (which is intentionally a bit shorter), which was just about OK. A few months later I started a new game in Stygian mode. I've reached week 32; I have 16 level 5-6 characters, but I've realised I need around 160,000 gold to upgrade all of their weapons and armour to level 5 (all guild and blacksmith upgrades are done). This is probably at least 8 weeks more before starting the endgame. I've given up for the moment; it was just feeling too grindy. Strangely, my Radiant playthrough actually took longer, at least in terms of weeks; I finished in week 56. But then, I had a few deaths, and wasn't playing quite as optimally in general.
Beyond this point, I'm going to continue to criticise (it's much more fun to point out the flaws in something than to sing its praises). But I should perhaps re-stress that I do think Darkest Dungeon is a good game; I wouldn't have played through it twice if I hadn't.
There are quite a number of what I would call “balance” problems. Curiously, this doesn't include balance between adventurer classes. I'm sure there are players with enough experience to be able to say that some classes are over or under-powered, but I don't consider myself one of them. While I've certainly formed some opinions, I can't make any confident statements about the power of any class. The dialogue I've seen in forums about this never really seems to reach a consensus from what I can see, though I appreciate that patches have changed the situation from time to time. I'm much more confident about some of my strategic assessments.
Where to start? Well, the eight buildings you can upgrade are not very well balanced. You need some Stagecoach roster upgrades, and fairly soon, too. One or two network upgrades are probably useful, too, though you certainly don't want them all. You pretty much need to upgrade the Guild and Blacksmith fully, and as soon as possible. You want a few Sanatorium upgrades, but only in the late game when diseases start to be common and only for disease treatment. One or two Survivalist upgrades might be useful (I've never gone for them, though). The Abbey and Tavern upgrades are entirely useless. The Nomad Wagon I believe is too expensive to warrant using. As I note above, money has been the biggest constraining factor for me in how quickly I can finish the game; XP and heirlooms come quicker than money.
On a related note, adventurer traits appear at first glance to be fundamental to the gameplay. But I've actually concluded that for the most part they really don't matter that much. Very few negative traits are a real problem at all (most of the “curiosity”-type traits will almost never impact negatively on you, for example. Kleptomaniac is the worst of this group, but it's really not a big problem). Even the best positive traits are fairly marginal; certainly worse than a single armour or weapon upgrade.
And there is really no good strategy available for even trying to manage traits. Removing locked negative traits is far too expensive to consider. Removing unlocked ones is still quite expensive, and the hero will likely acquire another one shortly anyway, so why bother? Locking in positive traits is also too expensive to justify. Dismissing heroes who acquire bad negative traits is far too costly in terms of XP lost. So mostly you just ignore them.
There's a curious lack of balance in terms of time spent at each of the three dungeon levels. In both my runs, nearly 70% of my runs were in Veteran dungeons. You gravitate quite quickly from Apprentice to Veteran because the increase in XP is so large, Veteran is really not that hard, and the XP gap from level 0 to 1 is so small. You're required to do Veteran dungeons from levels 3 to 5/6 anyway. But there's really very little reason to do Champion dungeons at all. The difficulty step up is huge. By the time you can possibly consider attempting them (you really don't want to be risking level 4 heroes in Champion dungeons unless you're a much better player than me) you're so close to the endgame that there's very little need to get any more XP on the hero. Resolve level 6, from what I can see offers +10% to all resists and absolutely nothing else, so it's not at all important.