Class balance analysis: Grave Robber

[geepope]

The much-maligned Grave Robber is a strong contender for worst class in the game. To be fair, she is much more consistent than many of the other weak classes. Classes like Plague Doctor, Houndmaster, and even sometimes Jester can find themselves completely sidelined in the wrong situations, whereas the Grave Robber can keep plugging away with her barely acceptable damage all day long. The problem is that it is essentially impossible for her to actually excel at anything. The Plague Doctor’s turns veer wildly in usefulness from “shut down half the enemy party” to “heal someone for 1 HP because it’s literally the only relevant thing you could do”, whereas the Grave Robber just manages to not quite do enough damage every turn. Worse, she doesn’t have a single attack good enough to lean on so she winds up needing a full slate of skills just to consistently do mediocre damage. She does have two things going for her: she’s one of the few fully back row capable characters in the game, and she’s pretty great at traps and scouting. But her back row abilities have been largely obsoleted by the Arbalest who brings better damage dealing abilities as well as support, and traps/scouting can be handled by much stronger classes almost as well.

Pick to the Face: Borderline to solid. It’s a standard attack, but it’s got a lot of flexibility and good acc/crit.

Lunge: Solid. A fun, useful attack. Not amazing, but it deals solid damage with a great crit rate and being able to juggle party formation on the fly has its uses. It scales poorly, though, and even at its best is barely what a stronger class can do with their basic attack.

Flashing Daggers: Solid. One of the weakest pure damage AOEs in the game, but AOEs are just that good. Works fine from the back row, too, and would be vaguely competitive with the Arbalest if she didn’t have better buffs and trinkets available.

Shadow Fade: Solid. Stuns have plenty of uses and getting a self-buff and movement out of it is a pretty nice little package.

Thrown Dagger: Borderline to solid. Basically a repeat of Pick to the Face, but for ranged Grave Robbers. With the crit bonus and self buff (tiny though it is) this is arguably a superior basic attack, but doesn’t work well with Lunge abuse.

Poison Dart: Bad to borderline. It has a niche for letting ranged Grave Robbers attack the front row, but it’s still one of the weakest DOT attacks in the game--which is to say, one of the weakest attacks in the game period. Generally doesn’t stand up to Lunge or Flashing Daggers for DPS, but can be almost decent against bosses and high prot enemies.

Toxin Trickery: Bad. Cures are handy, but the buff is smaller than the already-minor boost on Shadow Fade the massive. The speed bonus is deceptive: she’s already the fastest class in the game, so while a minor boost to beat out fast enemies consistently is nice +9 is overkill and isn’t going to let you do much in the way of turn order shenanigans. Good luck even fitting this into her skill list, though.

Snuff Box: Solid. When you’re as bad as the Grave Robber it doesn’t take much to cripple her, so being able to shrug off disease this easily is critical.
Gallows Humor: Borderline. Looking at the numbers it’s much better than I had given it credit on the Highwayman write-up. Recovery is still not my favorite use of respite, though, and when you do need it consistency is pretty critical.
Night Steps: Solid. Scouting always helps and it doesn’t take much respite.
Pilfer: Bad to borderline. It’s… kind of useful, in a unique way, and doesn’t take much respite. But it’s terribly unreliable considering there’s no guarantee you get anything usable. If it gave you several items or let you pick an item on demand it would be a lot more useful. Or hell, throw in the food from her deprecated Forage skill. 4 food + a random tool would be dandy for 2 respite.

Lucky Talisman: Baaaad. The Grave Robber is built as an attacking class and desperately needs every bit of damage she can scrape together. It’s not like she has support skills she can use this to spam better.
Stunning Satchel: Solid...ish, I guess? The Grave Robber’s stun is one of her better abilities, and she doesn’t mind being shuffled since she has so many movement abilities.
Quickening Satchel: Borderline. The Grave Robber has accuracy to spare and speed is nice, but she’s already going first most of the time (not that it matters, since she struggles to kill things on her own) and doesn’t really care about move resist.
Seer’s Satchel: Borderline to solid. The Grave Robber’s scouting is one of the few things she brings to the table. The penalty is pretty minor, she’ll still match the fastest characters for speed and it’s not like her turn is critical anyhow.
Raider’s Talisman: Borderline to solid. Again, the Grave Robber’s out of combat contributions are arguably more important than her combat contributions, and with the crit bonus this is arguably is a net bonus to her combat performance too. The hunger is actually the worst part, carrying extra food makes long dungeons a pain.

Builds: The Grave Robber has almost no support skills. She’s built as an attacker with options to go primarily melee, ranged, or mixed, with a niche off-role as dodge tank/controller. Her skills have solid synergy and distinct positional requirements and considerations, so there’s theoretically a clear distinction between a melee-primary Grave Robber and a ranged-primary Grave Robber. Unfortunately there’s not much in the way of buffs or trinkets to favor either ranged or melee attacks, and her damage is so low that having both Flashing Daggers and Lunge is basically mandatory. So in the end it doesn’t really make that much difference.

Suggestions:
-Wouldn’t hurt to bump up her base damage a little bit. Her skill spread is heavily biased towards straight attacks but she’s not very good at it. You might be able to get a support class to work with 4-7 base damage but the Grave Robber has very little in the way of support.
-It would also help a little to have Shadow Fade and Lunge either be both move 1 or both move 2, to make it easier to shuffle back and forth. The Grave Robber also badly suffers from only having one or the other available at any given time; it would be nice if Lunge was extended to row 2 or Shadow Fade was extended to row 3. Maybe even both; neither of them is really good enough to justify being as limited as they are.
-If you’re not going to let Poison Dart do real damage it should at least have a more relevant debuff--if it debuffed Dodge rather than (or in addition to) blight resist it would still help land blight attacks, while also helping land everything else. Maybe even throw in an accuracy or damage debuff on the target, since if you’re going to waste time on attacks that take multiple rounds to add up to any appreciable damage, you should at least get a better margin of survivability for those rounds. But she’s an attacking class first and foremost and I’d just as soon see her do full damage + full blight, to make darts more competitive with Lunge/Flashing Daggers.
-Toxin Trickery is an incredibly niche skill. Let her use it on teammates and give her a minor support role, or turn it into a real self-heal to make her into a survivable dodge tank. Or give it some damage buffs, so that at least in longer fights it could make up for the lost turn (having Toxin Trickery add blight damage to the Grave Robber’s vanilla attacks in addition to the cure/dodge buff would be cool, unique, and thematic, but not really supported by the engine at present.)
-For a more dramatic change, swap Toxin Trickery from cure + tiny buff into something like a self-inflicted blight + monstrously huge buff (like almost doggie treat level) and let the Grave Robber go Jekyll and Hyde on the monsters.
-Making traps progressively harder to disarm would make the Grave Robber’s skillmonkey niche more relevant. Currently everyone winds up with 100%+ trap disarm chance and the traps in champion dungeons do not seem noticeably harder to disarm.
-Trinkets need work. Would really like to see a row 4 damage bonus trinket in particular; Grave Robbers tend to move around a lot and don’t naturally hang out in back, so it would be an interesting trade-off.

[Remachinate]

Maybe it's semantics, but I don't see that as true. Even if the GR is worse than other classes, she's still capable enough to clear dungeons. It's not that she's an inadequate choice to beat the game (which is what I would consider "bad"), she's just non-optional. That's still an issue that should be addressed, but the context of the dungeons is just as important as the context of the other classes.

[h0b0king]

引用自 Baywatch：

Also, I'm not sure what you mean by fiddling. She's the most fiddly character in the line up if you want her to even come close to pulling her own weight.

Despite all that, I also enjoy using a Grave Robber but, the math doesn't lie. She's terrible at the moment.

What I mean is that while it's obvious Hellion etc are far better at doing damage then the GR I feel like I'm almost never healing, unblighting, repositioning or otherwise doing anything to keep my GR active. She just lunges, picks and shadowfades about.

With Highwaymen and Hellions they are always getting smacked, bled or pushed out of place and corrective measures need to be emplyed. Now I bet some of this is just perception, but I dio think she is hit far less then the other squishy damage dealing classes.

[Lampros]

引用自 h0b0king：

What I mean is that while it's obvious Hellion etc are far better at doing damage then the GR I feel like I'm almost never healing, unblighting, repositioning or otherwise doing anything to keep my GR active. She just lunges, picks and shadowfades about.

With Highwaymen and Hellions they are always getting smacked, bled or pushed out of place and corrective measures need to be emplyed. Now I bet some of this is just perception, but I dio think she is hit far less then the other squishy damage dealing classes.

Your perception is not wrong. In fact, the GR likely has the best avoidance of any hero. The question is how relevant that may be. Given her extreme lack of DPS, it is likely that on a cost-benefit analysis, her monumental lack of DPS likely outweighs her superior avoidance. At the moment, this problem does not become crippling except in a few DPS race-type of boss encounters. But then, in non-boss runs, basically any type of party configuration works if you are a good player.

[h0b0king]

引用自 Lampros：

The question is how relevant that may be. Given her extreme lack of DPS, it is likely that on a cost-benefit analysis, her monumental lack of DPS likely outweighs her superior avoidance.

I don't have the exact numbers at hand but is the Graverobber's lack of DPS really 'monumental'? She does something like 5-8 to start while I think the Hellion does 6-12? Yeah the top of that is considerably lower but the grave robber is usually doing like 6 or 7 damage? The Hellion is doing 8 or more likely 9. (ok 6.5 vs. 8.75 I guess). That doesn't feel huge given that the Hellion is a dedicated damage dealer far more then the GR. I think some variation is okay - I mean I think the GR deserves a break, but she need not be a total ninja assassin. I am sure there are plenty of other factors and that the difference gets worse as the heroes level but those raw numbers don't feel 'that' bad - especially with the crit chance.

[Lampros]

引用自 h0b0king：

I don't have the exact numbers at hand but is the Graverobber's lack of DPS really 'monumental'? She does something like 5-8 to start while I think the Hellion does 6-12? Yeah the top of that is considerably lower but the grave robber is usually doing like 6 or 7 damage?

You have to consider various damage adds (Quirks, gear, self-buffs - which the GR has none in terms of DPS enhancers), not just the floorline damage.

Further, a heavy AoE orientation is the most efficient DPS party configuration, and here the GR lags very badly.

[geepope]

引用自 h0b0king：

I don't have the exact numbers at hand but is the Graverobber's lack of DPS really 'monumental'? She does something like 5-8 to start while I think the Hellion does 6-12? Yeah the top of that is considerably lower but the grave robber is usually doing like 6 or 7 damage?

Grave Robber's base damage is 4-7, on a level with most support classes. Even with Lunge she can't match the damage on the Hellion's weakest attacks (realistically the Hellion is more likely doing 6-12 + bleed or 3-6 x3, and also scales much better with damage buffs.)

The Grave Robber doesn't have any more innate dodge than the Highwayman or even the Hellion (!!!) She does have Shadow Fade to boost her dodge while still contributing, but tossing out a stun won't always be the optimal move. Putting a GR in front and running into a fight where you'd really rather have Lunge than Shadow Fade sucks, and vice versa if you stick her in back. Adjusting position is easier for her than most classes but spending multiple turns to set up an average-at-best attack or stun is always going to be painful.

The GR is nice in that you can drop her into any slot in any party and she will be able to contribute something--just not very much, ever.

[Lampros]

引用自 h0b0king：

I am sure there are plenty of other factors and that the difference gets worse as the heroes level but those raw numbers don't feel 'that' bad - especially with the crit chance.

I forgot to address this directly, but the Hellion has higher crit chance on her signature single target attack than the GR in practice. The Hellion gets a 10 percent crit buff from camps, and camp self-buffs really change things. (This is where the BH falls behind the Hellion, the Arbalest, and the Highwayman as well, without a Mark.)

[h0b0king]

引用自 Lampros：

Further, a heavy AoE orientation is the most efficient DPS party configuration, and here the GR lags very badly.

Yeah I don't play with much AOE, my Hellions don't even have the skill - though I do use bolo, the jester's double slice and grapeshot occssionally and the Leper and crusader are always cleaving. I know it works, played around with it several months ago and found it sort of boring, so this is purely a personal choice - but I figure the game is more interesting without AOE cheese (though a Highwayman with +100% or more damage blasting down waves of foes is sometimes pretty cool). I won't disagree that the GR is not up to speed with other attack classes, but I don't think she's quite as bad as the consensus around here. I'd really like to see changes along the lines Geopope proposed, but I hope that any changes won't completely remake the class into some kind of murder ninja.

[Lampros]

引用自 geepope：

Even with Lunge she can't match the damage on the Hellion's weakest attacks (realistically the Hellion is more likely doing 6-12 + bleed or 3-6 x3, and also scales much better with damage buffs.)

Precisely. A truly efficient DPS team will have at least 2 - and likely 3 - AoE-ers, and they will be AoE-ing things down in most situations, not single targeting.

[Lampros]

引用自 h0b0king：

I won't disagree that the GR is not up to speed with other attack classes, but I don't think she's quite as bad as the consensus around here. I'd really like to see changes along the lines Geopope proposed, but I hope that any changes won't completely remake the class into some kind of murder ninja.

How "bad" the GR is really is a subjective issue; but I think objectively - at least by using DPS and utility metrics combined - it is not unreasonable to say that the GR is comfortably nestled in the worst tier. I'd say only the Leper is in competition as a worse class, with even the Hound Master being better, due to the far superior utility he and his dog provide.

[Lampros]

As a general matter, I think a lot of folks are missing an important aspect of these balance discussions. At the moment, the trash mobs are so easy that any configuration can with ease enable you to breeze through any dungeon. So one may think a grossly under-powered class is fine, because you are looking at the under-powered class in relation to the monsters, not to other classes. But this may not be the case in the future; and even if it remains the case, that is not a legitimate way to do a class comparison. So the fact that one can run 4 GRs in a party and finish a dungeon is not really an argument for the class being balanced.

Edit: There is an exquisite parallel to a similarly wrong-headed discussion I found myself in a few weeks ago. One dev from another game tried to refute my view that an x class was grossly under-powered, because that class can solo entire dungeons (it's a game where the max party is 12!).

What he ignored was the fact that the same can be said of virtually any class in that game, and those other classes can do it FAR FASTER than this particular under-powered class x. The monsters were indeed that easy in that game.

[h0b0king]

引用自 Lampros：

As a general matter, I think a lot of folks are missing an important aspect of these balance discussions. At the moment, the trash mobs are so easy that any configuration can with ease enable you to breeze through any dungeon. So one may think a grossly under-powered class is fine, because you are looking at the under-powered class in relation to the monsters, not to other classes.

The question for me is does this mean that the Monsters should be buffed along with all the classes or that the best classes should be gimped a little bit? I suspect that were AOE far less powerful (I find my game is more fun now and harder - I've actually lost people without meaning to - but I'm playing with my own odd rules against certain tactics and strategies) we'd find that the regular monsters were more challenging.

I don't know that the metric for the game to be successful should be one of perfect class balance and maximum challenge requiring optimal tactical choices to survive. I think that might be a bit overwhelming for new players and a bit boring for the few that got deep into the game, because it would constrain playstyles that were sub-optimal. I know there are players that love optimization and efficency - but I suspect they are a minority. I'd love to see a game that had heroes who were good at things and bad at things, not necessarily in equal measure, but in a way that made various tactics viable. I note that having stuns is more useful now, as are DOTs, moves and hitters that can reach multiple positions (another reason AOE is very powerful). I'm pretty happy with the way Darkest Dungeon is reaching towards its balance - and winding between being harder (as it is after Corpse & Hound) and easier (as it was after the prior patch). I'm confident that it'll turn out pretty solid.

[Baywatch]

引用自 h0b0king：

I don't know that the metric for the game to be successful should be one of perfect class balance and maximum challenge requiring optimal tactical choices to survive. I think that might be a bit overwhelming for new players and a bit boring for the few that got deep into the game, because it would constrain playstyles that were sub-optimal. I know there are players that love optimization and efficency - but I suspect they are a minority. I'd love to see a game that had heroes who were good at things and bad at things, not necessarily in equal measure, but in a way that made various tactics viable...

I would hope a challenging game that requires you to play optimally with perfect class balance is the end result of all of these balance discussions. If it's done right there shouldn't be any sub-optimal playstyles. Right now, when you look at her kit at face value the Grave Robber doesn't have a playstyle aside from undewhelming dps character.

[Lampros]

引用自 h0b0king：

The question for me is does this mean that the Monsters should be buffed along with all the classes or that the best classes should be gimped a little bit? I suspect that were AOE far less powerful (I find my game is more fun now and harder - I've actually lost people without meaning to - but I'm playing with my own odd rules against certain tactics and strategies) we'd find that the regular monsters were more challenging.

This is the crux of the problem. I prefer bumping monsters and weaker classes, rather than nerfing the stronger classes, because taking away what exists is always more painful for the players who have grown attached to a particular class or style of play. But, on the flip-side, from a dev perspective, it's easier to nerf a minority of classes than boost a majority of classes/monsters. So you almost always see the latter balancing solution in most games.

As for your point about nerfing AoE in particular; the issue isn't really AoE stacking as much as DPS stacking. Here's the fundamental design problem. In most games, a high DPS approach is well balanced between risks and rewards - or costs and benefits. Typically, a high DPS approach will enable you to finish more quickly, but it will also make you significantly more vulnerable to being wiped. So there is a balance between a DPS approach, and a non-DPS approach. But in DD, due to the stress mechanic, the balance shifts dramatically in favor of the DPS approach, because the greater risk of wipe in this game comes from stress accumulation, and not health loss. So unless you slow down stress accumulation or greatly increase ways to counteract it, you are already forced to go the DPS approach.

So the fundamental game mechanic - stress - already forces you into a monist approach.

[Catgirl Stuffed With C4]

引用自 Lampros：

As for your point about nerfing AoE in particular; the issue isn't really AoE stacking as much as DPS stacking.

Well, that and buffs add in straight addition which completely invalidates damage penalties that are supposed to be the reason not to use it as a one size fits all attack. AOE just exemplify that.

You can boost literally any attack to having a positive damage modifer, just in most cases this leaves most classes main attack still being the strongest. AOE on the other hand, end doing more net damage - since any damage overflow when you OHK a single target really doesn't matter.

Like for example, Poison Dart - an otherwise universally agreed bad attack in GR's kit.

It has a -90% modifier right? So, you get a buff - camping, quirk, trinket, whatever - you get 20% more damage.

Except. You're not actually getting 20% more damage in this case, your erasing part of the innate penalty for the attack, which is MUCH more than a 20% damage boost.

Instead of a not even registerable increase from the 0-1 damage because the attack is way too low to raised to a noticable degrees by an actual 20% damage increase. Instead - since you're wiping away the modifier by 20% you end up shooting from 1 damage to 4 - which is a 400% buff to the damage of the attack.

The problem only gets worse too - since ironically only the already hard hitting heroes are the ones with huge self-buffs, making it an mad dash to just kinda slosh all the damage unto the person with the biggest AOE and reaping the benefits.

All because buffs are cumalitive instead of multiplicative in nature. This would quickly curb the crazy damage ramping on mutlitarget skills since you wouldn't ever be able to reach that default base damage state on tacks with negative modifiers.

Let's take Arbalest now - you buff her up Restring Bow and have a Wrathful Bandana on her.
Man, Sniper Shot sure is packing a punch now already! You fire bolas, huh, that's weird it's doing the same damage as Sniper Shot before I camped but to two people.

Then you end up with ridiculous cases where you - throw on another +20 Damage trinket and decide to go in a party of Arb/Occ/BH/MAA.

Wrathful Bandana and Legendary Bracer having already brought Suppressing up to Arbalest base damage.

You camp.

You use Restring Bow +20 DMG and +5% Crit.
You have Occultist cast Dark Strength +20 DMG.
MAA cast Instruction giving her another +20 DMG

All-in-all the transaction nets you +100%

This is where the difference would come in - and to illustrate, let's say the base damage for Arbalest is a straight easy 10. She scores 10 on Sniper shot all day, everyday. Suppressing Fire on the other hand hits 3 people for 2 damage each. This is great, because SF is netting less damage, ensuring you want to continue to use SS.

However, due to buffs, SS is now hitting for a solid 20, doing indeed 200% of its original damage.

In a system where buffs weren't added cumatively Suppressing Fire would see a pleasant increase as well - doing 200% of the attack's damage. Which would be 4 in this case. This allows both skills to coexist happily - your single target attack generally being preferred.

Yet - in the current system that -80% modifier SF started with is gone and even replaced by being 20% higher than the base damage. So - instead of jumping from 2 to 4 it jumps from 2 to 12. Damage against each target it hits being effectively 600% higher that the attack originally could do. This especially becomes problematic - because suddenly the AOE exceeds DPS of your single target attack.

So, yes, Stress encourages DPS stacking - but ultimately the way the current system is really the problem for why AOEs and damage scaling in general are completely broken. I even have a screenshot off steam (couldn't figure out steam screenshots on mac) of an Occultist using Weakening Curse for the exact same (sometimes higher) damage than an unbuffed Man-At-Arms using Crush and that just shouldn't be possible.

Anyhow - I know a lot of balance talks grind to a halt when people mention numbers, but it seems like a problem that everyone chooses to ignore.