Data for Rebalancing DPS

A lot of discussion has been had regarding the gap between DPS classes, and calls have been made for ages about rebalancing the classes. The devs have stated that rebalancing of the classes should be coming in Q4. This post is here to try and help with the rebalancing.

I know the devs don’t play the game at the same level the players do, so I took the time to gather some of the end-game players to run DPS checks so I could get an idea of what each class is pulling at the highest level of skill.

Method:
I had players that were reputable in their class assist me with this. We would go into the Sewer dungeons as a pair: I would play Warrior, and the other player would play the DPS class they were playing. I would spam Provoke on Mutated Rat, and the other player would kill the Rat as fast as they could. Tiles and affixes were fine as those are tied to the class armor. External buffs like potions and other class damage boosts (Affliction Weakness stacks for non-mages, Shield Bash, etc) were not used, as I wanted class-specific DPS. I had each player run this five times to get an average for them. I had five players in each class. This lead to one hundred total runs. I need a shoulder massage because my goodness Provoke.


What we see is that Runemage is far and away the top DPS class, followed by Shaman, Scoundrel, then Ranger, in order. This is something that the community has known for quite some time, but this time instead of it being just “common knowledge”, we have definitive numbers in a controlled environment showing the results.

Class by class:

Mages were top DPS. This is where it should be due to the skill requirement to play this class effectively.

Currently, the end-game Rangers are pulling less than half of end-game Mages (average ~48%). While Runemages should be top DPS due to the skill requirement, this gap is hideous. It’s no wonder that only a handful of players running end-game content play Ranger as their main class. Even though it is a really fun class (as far as fun goes, my favorite damage class), you’d need two Rangers to pull equivalent Mage numbers. Why would I bother when trying to clear hard content that needs high numbers?

The Scoundrels on average also pulled around half the Mage numbers (average 50.4%). There were much more variable numbers here, though, showing that the Scoundrel rotation is still a puzzle to be solved for many players. Even so, the top Scoundrel single run was still only about 70% of the average Mage numbers, and 60% of the top Mage run.

Shaman numbers were fairly consistent between runs of the same player with one exception. Shaman #1 ran the first two runs at a distance from the Mutated Rat as to avoid poison. The next three runs the Shaman stood next to the Rat, ignoring the poison damage. While this did drop the average DPS, I think it makes a good point showing that Shaman numbers drop significantly as they get further away from the target. Even so, Shaman average was 69% (nice) of Mage average.

When it comes to fixing the gap between the DPS classes, I think a good step forward would be a flat 66% increase to physical DPS. This would bring the Ranger average to ~60k (coming in at around 80% of Mage DPS) and Scoundrel average to ~63k (coming in at around 84% of Mage DPS).


On top of just fixing the gap for boss damage, both Scoundrel and Ranger are range-based single-target damage which sucks for trash mobs. Scoundrel needs curved shots which are single-target only, and Ranger needs range which is difficult in many dungeons (not to mention weak points being single target specific). Shaman and Runemage can sit right on top of the enemies and be just as, if not more, effective as being at range.

I think Shaman is right where it needs to be. Due to the extremely low skill ceiling for the class and simple rotations, it’s an easy DPS class. Having decent DPS with low effort is offset by the fact that any fight with distance (Mist Keeper, for instance) or movement (Dovregubben, Sewer Slime) essentially wrecks a Shaman’s numbers. That and the persistent orb not spawning bug, but at this point I’ve accepted it as a part of the class. Probably have to thank Rupert more.

I don’t think Mage numbers should be reduced at all. The last time Mage was nerfed (Triplicity), many mages left the game and haven’t returned (not that I blame you for nerfing it, Trip was just a lotta OP). However, beginner to mid-level mages struggle to pull their weight, and a nerf to Runemage will bring them down even further.

TLDR: I’m pushing for a 66% increase to physical damage, and no nerf to any class in order to bring things closer in line. Supported with player driven data.

26 Likes

Really nice work! Agree with the conclusion.

3 Likes

appreciate the data, I also would love a ranger buff

4 Likes

this is outstanding stuff! i really hope devs use some of this data :crossed_fingers:

1 Like

A note on the comparison, crit chance increase (by drinking a crit potion) makes the shaman and scoundrel dps scale differently than the mage and ranger.

Scoundrel will be a bit weaker % wise in comparison to the mage/ranger boost but will still be higher dps than ranger. Shaman will do around the same % boost but differently than ranger/mage. Because it starts with guaranteed crits those are out of the equation making it less useful to increase crit chance. Luckily lightning critting makes it respawn, which fights back that % loss to about equal.

Another note on these numbers, they are respresentative on the players and their dps on that day, but they are not representative for max performance for those classes. It does show how skilled the end end game community is in their class type for sure.

6 Likes

Meeeshka good work!
Worth noting is that the strongest classes will attract the best players, as optimisation is not only about picking the strongest talents, but also picking the strongest classes.

2 Likes

While that’s true, it’s also true about any other boosters like Int potions etc… I assume that in the experiment none of the participants took any pots and all the classes still had their base effects and would’ve still had a level of Int and Crit appropriate to their class applied to damage. That said, it would’ve been good to record their stats at the time for consideration.

not representative for max performance for those classes

I don’t know why that really matters here since the numbers are averaged across 5 people in each class and they’re pretty consistent. Additionally, for the exception of ranger Rapid Fire and mage casting, which depend on the speed and accuracy of the player, all the other max DPS for each class is limited by the pacing of the mechanics. Rapid Fire was severely crippled and doesn’t compare anymore to the Precision (limited by pacing).

I’ve got all the Int/Str scores recorded on a separate sheet, but even after taking those into account, it didn’t make any significant difference. No potions were used in these runs, although I plan on doing this all again with full potions after my shoulder recovers from all those provokes!

2 Likes

Intellect and Strength if both are used will give all 4 dps classes an equal boost. Some only need intellect to boost their dps, others need strength and Intellect. So those don’t need to be noted for the test between classes. Because they have no effect % wise for the difference between the 4 classes. They both boost damage types by 8.33% (best potion version which is 250 points = 8.33%). So if you are a mage your Int potion will boost all your spells by 8.33%. If you are a scoundrel Int will only boost your cards damage by 8.33%. But drink a strength pot only and you will get 8.33% for your normal bullets, but your cards don’t get 8.33%. Together Intellect + Strength only 8.33% boost for scoundrel. Which is the same as mage.

Not really consistent. I know the people who ran the tests were failing on their class dps on some tries because of a bug (shaman totems breaking), or just failure in general. I know this because I myself made big mistakes (or bugs) multiple times on multiple runs that were recorded. Which sometimes doesn’t happen in a fight for some players. Depending on how good they are and how they feel that day.

@Meeeeeeeeeeeeeeeshka I think he meant things like armor re-roll of crit chance + projectile damage + used rings? Because they can also make a difference together with weapon and weapon affixes.

3 Likes

Yes, i understand that and agree with you, but all classes also have criticals even unboosted. However, mages get more advantage from the boost to their consistent damage not relying on lucky swings. I don’t play shaman or scoundrel so you can correct me if i’m wrong but i imagine that even with a full crit loadout and critical pot it would still be really difficult to have every crit deal maximum possible damage. So in the duration of the full fight, it would average out and still not come close to covering the 30%-50% difference we’re seeing displayed above.

Not really consistent. I know the people who ran the tests were failing on their class dps on some tries because of a bug (shaman totems breaking), or just failure in general.

Yes but all the classes were tested in the similar time frame… don’t you think the other people had similar experiences? I mean of course the data would be much more interesting if we had a much broader population size but i don’t think Mishka could handle it :smiley:

I think he meant things like armor re-roll of crit chance + projectile damage + used rings

Yes, sorry if that was unclear, i meant all the stats shown in the player panel as well as armor/weapon damage/crit affixes

2 Likes

I am guessing you are talking about the sad truth that ranger has one out of X arrows being a way bigger hit (the charged shot), so luck plays a bigger role in critting that hit. Ya, but other then the luck of bleed on that hit, crit damage luck will not make a huge difference. Mages do not get more advantage by hitting more often. Their affect of lucky crits is less noticeable on their dps from fight to fight. But luckily they have plenty of other random crap like how perfect a single spell is cast that affects their dps fight by fight.

I just did some checks on my dps calculator and these are the extremes in dps between getting crits on every charged shot and not getting crits at all:
Average: 100% (not giving away numbers now xD with the charged strikes potions and max crit chance + max crit damage possible of the ranger)
No charged shots crit: 93% (worse case scenario ever)
All charged shots crit: 115% (Best scenario ever…)

These extremes average out to the 100% mark how longer the fights are. On the scale on a whole shard it is probably always around 100%. Not more, not less. And a single boss fight might typically fluctuate a few percent up or down the 100% mark.

2 Likes

Interesting. I would be really curious to take a look at that calculator privately if at all possible :wink:

Anyway, thank you for the good discussion I do think that there’s definitely balancing that needs to happen in a lot of the classes and hopefully this will drive some of that forward :man_shrugging:

As far as ranger boosting, I personally think that instead of a flat 66% boost across the board for ranger specifically it should be boosted scaled to distance. Meaning that at 0m the current damage would only increase by say 30%, at 20m it would increase 66% and at 40m it would increase by 90% (in addition to existing distance modifiers).

And we none of us want this to ever happen again :smiley:

Mmmmmmh, maybe. I do think you should not reward any ranger past 30 meters though. Because alot of fights, shard affixes and healer skills really force the ranger to stand closer. If not done gives everyone trouble. Not the ranger only. And if you would get more dps by shooting for the stars (50+ meters away) there will always be people doing that xD. But again you would make the whole party angry and depending on the dungeon/fight not even possible to do that. Making ranger even more situational specific handicapped. I think the 30 meter talents are already accomplishing what you want, to be honest :sweat_smile:

2 Likes

haha yeah that’s a very good point :smiley: change that to 30m

I’d like to thank you for putting this stuff together and everyone who contributed to the undertaking, it’s always super helpful to get more data in everybody’s hands (including ours) and we appreciate it.

I can’t speak on how this would or will impact any balancing decisions going forward, but it’s certainly better to have practical numbers and the more players involved the better it definitely is to get the accuracy of them and something we can evaluate in an easier way.

11 Likes

Didn’t realize I wasn’t going to be the only one without a +6 for the tests but feel like I did well for my runs.

3 Likes

While this topic is up, I’d like to make a comment on how broken mage is in pvp. While fireball spam is absolutely ridiculous blocking entrances to caves, there’s an even bigger problem in that it only takes 2 greater afflictions to kill a player. I see far too many mages in battlegrounds that land a greater affliction, hide, then land another one. It’s impossible to fight as it’s basically a strong poison effect that lowers your defense. Considering mages control where their spells land, it’s hard to dodge a stream of fireballs or a greater affliction. I don’t mind if they keep their dps for dungeons, but something needs to be done about combat. I’d suggest a magical defense on gear alongside the normal armor stat.

2 Likes

remember, gear doesnt matter in battlegrounds.

Right it’s been a while since I’ve battled because I don’t like mage dominated battlegrounds so I’ve forgotten a lot… perhaps a built in resistance to magic then. Anything to balance mage pvp.

Mages are incredibly easy to kill in pvp
They are weak and only few are good short range
All I can say is get good ig

3 Likes