Balance Megathread

[Ceno]

DISCLAIMERS:

1. At the time of this writing, I can’t actually play GD. Any comments I make here are with respect to former builds (v1.0.0.1) and from feedback I’ve received from the (many) places I’ve got my ears near.

2. I’ll be referencing things from the Game Info and Features subforum, maintained by eisprinzessin. eisprinzessin writes of the subforum the following:

Most threads in this forum have not been updated in [a] long time, but a lot of information is ageless.

So perhaps intentions have changed. Without further official comment on the matter, I’m going to assume that - with respect to balancing discussions - intentions have, however, thus remained unchanged and that anything I quote is essentially gospel in nature.

3. As some of you have probably noticed by now, adoomgod and I (with possible input from others) have been looking into a rebalancing mod. This post is not intended to get your input on that or to take away responsibility or recognition from Crate. The rebalancing mod is an attempt to approach Grim Dawn from a different direction. This post, however, is an attempt to collect people’s thoughts on the present direction of Grim Dawn, as being developed by Crate, and we should recognize Crate’s efforts in that matter as such. In actuality, I haven’t consulted with adoomgod once about the topics in this post, this is solely my writing and initiative.

END DISCLAIMERS

One of Crate’s biggest focuses has been on improving the feel of combat and pacing so as to better capture the frenzied killing that is core to ARPGs. [1]

Death comes a little quicker in Grim Dawn. Combat is balanced so that you can basically be killed a little more quickly if you don’t watch out. If you get surrounded, for example, you’ll die a lot faster than in Titan Quest, so tactical movement is more important. There are also some enemies with pretty hardcore ranged attacks that you need to keep an eye out for and avoid. The developers also created a number of enemies with high-damage attacks preceded by special animations, so if you’re paying attention, you can quickly get out of the way. More effort was put into creating encounters where players have to think a bit and keep moving around. [2]

In an ARPG you’re fighting hoards of enemies and the addition of more enemies doesn’t really matter much unless you’re pinned in a corner or something. [3] Most enemies are designed as fodder and you can typically kill them in 1-2 hits anyway. They are only meant to become a threat in numbers. [4, 5]

Fodder enemies dying in 1-2 hits doesn’t necessarily make the game easier. If enemies can survive longer, potentially getting in more hits on the player, then each hit needs to do less damage. Generally this makes for slower paced, less threatening combat. Either that or you need to restrict the number of enemies that can be attacking at once. People enjoy fighting large waves of enemies, so the developers are balancing the game in a direction that supports this.

They want to create a combat dynamic more typical of a zombie game where it is about keeping the enemy off you as a ranged class or, as melee, making sure you’re not surrounded by too many foes at once. Basically, you want to be killing fodder creatures faster than they can accumulate on you. Unless your character is a real tank, your health will start to drop rather quickly once you get more than 2-3 fodder enemies attacking you at a time. [6]

The first thing that needs to be done is that damage should be less consistent. Average damage caused by enemies should be lowered so that the player isn’t just constantly hemorrhaging out health. [7] In ARPGs the sense of danger comes more in bursts where you’re suddenly threatened to be overwhelmed by the amount of damage you’re taking vs. the amount of damage you can quickly regenerate.

The immediate threat could also come from enemy skills that could quickly kill a player if they landed in succession or while your health was already low. [3] But if fodder monsters all had the potential to use a powerful skill, it would make encounters very difficult to balance since there is currently no way to prevent them from all randomly doing it at once. [4]

There should be more pronounced spikes in damage that give you those oh-shit moments in combat. One of the ways this will be accomplished is with fodder enemies that miss the player more often and have a wider damage range. [7]

By “most enemies” the developers mean most of the enemies on the screen at any given time, NOT most types of enemies. Since you’ll generally be fighting large numbers of fodder enemies at a time with only a few more powerful champions / heroes interspersed among them, by number, most of the enemies are fodder. However, in terms of types of enemies, there will probably be a greater number of champion types since it is those enemies that really can mix up the combat with skills that create unique situations. For example, one of the gazers (the floating brain type things) has a “frenzy” aura that increases the speed and regeneration of enemies around it. This makes its zombie minions much more dangerous and difficult to kill unless you hunt down and kill the gazer first and the gazers are good about eluding the player. The game won’t be too easy. [6]

[1] That’s for damn sure. Grim Dawn at its worst feels a helluva lot better in terms of combat than Titan Quest did at its best.

[2] Nothing mentioned here is explicitly wrong. I do, however, feel like it doesn’t paint the whole picture. But I think that’s more to do with [4] and [6, 7].

[3] The first of these two references doesn’t seem entirely true anymore. Perhaps this is because enemy movement speed is particularly on-par with that of players, and that it is fairly hard to pull certain enemies of a pack away from the rest of the pack? This seems more like an enemy balance issue than a player one, as giving the player more movespeed would only let them create more of a gap between themselves and enemies, whereas what we need is the ability to pull more dangerous targets out of the ranks of less dangerous ones so that they can be handled without being swarmed. I think some degree of this was attempted at various points of development - Reanimators, for instance are far slower than every other Aetherial enemy - but it doesn’t seem as though this has ever been a focus of balance nor has it been continually adopted to more modern enemies (like Chthonics) or humanoids. This goes for both melee and ranged play. Ranged players ought to be assaulted first and foremost by non-ranged high-danger targets, and melee players ought to be able to lure such targets out of mobpacks in order to eliminate them one-by-one. Then it becomes a matter of cleaning up the trashiest of the trash left over when this iterative process completes.

I don’t know what to read into the second [3] point.

[4, 5] What really is ‘fodder’ anymore? Common (white) mobs? It seems that the vast majority of Ultimate isn’t fodder anymore, which could potentially be an issue. We’re surrounded by various Champion (yellow) mobs. I for one like this in concept - because Champions have added traits or characteristics that change up how they attack a player or defend themselves. However, it feels as though Champions have, as of late, been considered to be ‘not fodder’ insofar as the definition above is concerned. Champions are dangerous even alone, and in Ultimate, we get a vast array of them. Its worth dwelling on. I don’t have a suggestion here one way or another, just recommending that we reflect upon Champions and their popularity relative to the design stated above.

[6] It seems this is no longer the case, which is a shame, because this concept would seem to work.

Grim Dawn Version 1.0.0.2

I feel another issue is that by ultimate, the %dmg bonuses at their current levels on devotions may not offer enough gain in clear speed. On account of the overly high enemy damage and possibly too low enemy health, it is much more advantageous to just dump everything into defense.

If enemy life was a little higher and player damage bonuses were raised, it would increase the value of investing in offense. Lowering enemy damage would reduce the requirement to all-out invest in defense. Then melee players would be able to keep more sustained contact with enemies and ranged players might feel a little more threatened in tight spaces if enemies don’t drop quite as fast.

As it stands, most Ultimate enemies are actually fairly tanky relative to an average character, FODDER INCLUDED. This is so much so to the point that bosses are literal meatshields. But oneshotting fodder is not a common occurrence either way.

Furthermore, it seems as though the latest medierra quote also seems to contradict the concept of fodder being especially dangerous once more than 2-3 of them are attacking you. This is a circular conundrum we’ve walked into, apparently. Ought fodder be dangerous enough so as to warrant concern? Probably. Unfortunately, its presently warranting so much concern that the concern is addressed in character build design rather than character gameplay, which is the heart of why we’ve been seeing so much linearization in build development. So I certainly agree that reducing the damage of fodder should be a priority so that builds can afford to be made less defensive, and instead play more defensive. But I’d also still like to see fodder retain its conceptual and practical danger such that some number of them hitting a player is a significant concern unto itself. I hope that makes sense.

Again, I do think these problems could each be addressed with enemy mobility changes. Take Killing Floor/2, for instance. In the KF series, fodder (Clots, Crawlers, Sirens, Bloats (despite their tankiness)) are generally very slow. On the other hand, however, high danger targets - Scrakes and Fleshpounds - run their targets down like there’s no tomorrow. I think this is a good concept to emulate for Grim Dawn. Seeing an immediate assault from high-threat enemies will - again - prompt and allow us to handle them before we move to everything else. Alongside, you know, creating a sense of urgency and danger among players. Dreadguards kinda do this, but they’re not entirely dangerous in reality. They’re an example of bursty fodder than is easily slain but charge for you right away.

[7] The first point here is - seemingly - what is presently trying to be fixed with Ultimate. Everything seems to be “hemorrhaging out health” from players constantly, and this does not seem to be the intention of design. Which brings us to the second point, a fantastic concept that I’m afraid is woefully forgotten in Ultimate Grim Dawn. I’ll requote it here because it’s such a good concept:

There should be more pronounced spikes in damage that give you those oh-shit moments in combat. One of the ways this will be accomplished is with fodder enemies that miss the player more often and have a wider damage range.

Anyone who has played in Ultimate knows that this is entirely not the case. Fodder will ALWAYS hit you. In fact, fodder tends to have a ~15%+ chance to crit you. Player DA tends to wind up around 1600-1900 toward max level, unless you’re making a really-tanky DA character (in which case you can push up above 2900 DA).

Come to think of it, I feel as though this has never been the case in Grim Dawn. Leveled fodder has always had a decent chance of hitting players. So the obvious solution here is to reduce enemy OA substantialy, aside from particular enemies which are made to hit/crit you a lot (Furies, for instance, or Hornets with their Queens giving them +25% OA). Bring most fodder out of the crit range of 1600-1900 DA character builds, which will, in turn, reduce their chance to hit players considerably. As it stands, not-being-hit by attacks is only achieved through % Chance to Dodge Melee Attacks/Projectiles. This is why Maiven’s/Blast Shield/Overguard are so favorable; they are a reaction to the reality that EVERYTHING will hit a player, and thus we’re working on mitigating the power of those hits. But if EVERYTHING does not necessarily hit a player, mitigation becomes less of an issue as well.

Having said all that, we do need to remember the second part of that quote as well. Fodder enemies should - by this intention - have substantial damage ranges so that when they do hit you they can actually hit you pretty hard but average out to hitting you mildly, with the potential for very weak hits as well. So think of the damage ranges on Maces, essentially. Whereas right now we’re probably working with Axe damage ranges, which are very tightened and don’t have exceptional variance to them.

---

There’s a lot more to gamebalance than just combat, but I’ll leave that discussion to you guys (as I can’t presently hunt down the new Treasure Troves/Nemeses). There’s things to consider like player/item skill design, loot droprates, etc. But right now combat seems a little flawed and unfun; at worst, its brutally and insurmountably difficult. At best, its a little disinteresting. Both sides of this coin should be corrected, and that was the intention of this post.

Moving forward with this thread, I’d appreciate it if you offered suggestions to the best of your ability when you mention something being out of whack. Are Treasure Troves overpowered? Say so, and say what you’d expect (or want) to get out of them in a balanced scenario. Balance will not be achieved by Crate team alone; there’s dozens of us that can help pull it together. Do so.

[Emanresu]

For raising the threat of trash, the first thing that popped in my head was to not only open up the damage range, but dynamically scale it based on how many enemies are attacking you. For example, an enemy by itself could do 10-100 damage to a character. But when other enemies are attacking the player (aka if OA/DA checks have been rolled vs the player in the last X msec), the damage range could be tightened to 40-100. You could also apply this mechanic to a one-time mob OA increase.

That would recapture the dire feeling of being swarmed without making each mob a potential threat by itself. Clearly lots of trial-and-error would be needed to settle on the exact formulas, but on a theoretical level it seems workable.

I’m unsure of how movement speed should be addressed. You could give all melee Champions a Blitz/Shadow Strike skill, but then all Champions would feel very similar. Playing with Threat generation could be viable, but I’m not familiar with how it operates. CC that only affects trash mobs, like OFF, already helps separate out the priority targets, but not every build has a straightforward way to do that.

Elite always feels like the sweet spot of player power vs mob strength. Once you hit Ultimate, you never quite make it back up to that sweet spot until you start snagging numerous BiS gear. Rather than a continuation of Elite difficulty scaling, it would be more interesting if Ultimate offered changes to the combat dynamics. Obviously smarter AI would be the ideal, but that’s hard to program. Perhaps the addition of Ultimate-exclusive Champions would fill the niche nicely, but again that requires more work than tweaking numbers. Fewer trash mobs, but more Champions and Heroes is a simplistic way to approach it, but then the player doesn’t feel as powerful since not as many mobs are being killed per minute.

[sfbistimg]

No Ceno No

I’ll pay you to concentrate on a Chain Lightning Class first

Especially if you use the animation from Storm Tap (just don’t let it hit so many targets as 12 is way to many for Spammable chain Lightning)

[Ceno]

(off-topic)

Especially if you use the animation from Storm Tap (just don’t let it hit so many targets as 12 is way to many for Spammable chain Lightning)

I was considering making it a channeled skill like AAR (no offhand req). But don’t know whether that’d be ridiculously broken.

[eralduspr]

I prefer a Lightning Ball spell like the one from TQ…

I love throwing my balls a the enemies’s faces.

runs away

[Paavo]

(off-topic)

I was considering making it a channeled skill like AAR (no offhand req). But don’t know whether that’d be ridiculously broken.

Don’t want to derail this right off the bat, but this is what i was thinking as well

[adoomgod]

Ahem

I suppose I should reply here, though it is unfortunate that a lot of what Ceno is talking about may have been completely turned on its head by part 2 of patch 1.0.0.2…

First of all, I don’t even like the term fodder. It’s too subjective despite the concept being clear.

I’ll be defining fodder as enemies that provide no feeling of threat or practical threat to average players. So the issue with definitions is some people might define fodder as low threat, and some people might make fodder not about threat but about white named monsters in general. Hmm.

So I’ll just be saying that my philosophy is, that when geared and built decently, some fodder (my definition, 0 threat) should still exist. But there should be less fodder on elite than normal, and less fodder on ultimate than elite. This is already done simply by scaling their stats upward, but it will be done more interestingly by adding new monster types, giving monsters new skills and tactics on elite and ultimate etc. New Mechanical difficulty is something I hope to mod in with or without Ceno after we do a rebalance mod together.

I think all the shades of grey should exist too. Some mobs that are threatening to some builds/play-styles but not others, some mobs that are minutely threatening to everyone, some mobs that are medium threat, some that are situationally threatening and so forth. Imagine I add a champion to the game that has a large aura that debuffs player pets dmg by 30% and their health by 40%, multiplicative. Maybe I’d add a yellow-named groble that does a lesser version of this. Mobs that nerf only cast speed, mobs that dramatically reduce your armor if they hit you, etc. There are a lot of dynamics that can turn fodder to non fodder.

As for mobs surrounding you, well that’s a threat because you’re just literally getting hit more, but after their dmg has been scaled down new interesting things open up. I can add a default auto attack skill to a type of mob that can stack on itself 5 times with bleeding dmg, and if you get surrounded each monster’s 5 stacks stack with each other’s 5 stacks. DAS A LOT OF STACKS. They’d be melee so it’d be simple enough to avoid, but if you stand still like a fool you will be bled out. This mechanic is more interesting because instead of just taking dmg for being surrounded, you take more dmg the longer you’re surrounded. This is just one example of how to make groups of enemies interesting. Gazers are already a good example of the kind of stuff I like, as their buff on their allies is definitely noticeable (haven’t checked how strong it is in the new patch though). Another example would be a group of 7 enemies that does more dmg the more of them are alive. So instead of dmg that ramps up, their dmg ramps down as you pick them off.

I just wanted to quote one thing, “This goes for both melee and ranged play. Ranged players ought to be assaulted first and foremost by non-ranged high-danger targets, and melee players ought to be able to lure such targets out of mobpacks in order to eliminate them one-by-one. Then it becomes a matter of cleaning up the trashiest of the trash left over when this iterative process completes.”

I don’t think this should be a general rule. I think it should apply in some cases (it doesn’t exist at all in the current meta basically), but other groups of mobs should be harder on melee characters by not allowing this kind of combat. If it isn’t clear, I am saying the addition of this sort of thing is welcome, and a goal of mine, but I don’t want it to apply everywhere. I want some enemies to piss on melee, some to piss on casters, some to piss on ranged auto attackers, some to piss on summoners, some to piss on tanks, and all the shades and combos in between.

I completely agree with Ceno about enemy fodder OA on ultimate. I want to make DA more relevant than it is, and the first place I’d do so is toning down fodder OA. But again, not in all places… just in most :]]]. Some fodder can do less dmg relative to other fodder, but have much higher OA. God I love diversity in the morning.