Thought Experiment: What is Viper's Weapon Damage Equivalence?

WhiteMage

Major TL;DR:
This page is incredibly long and less a presentation than a big scratch sheet of paper for me when doing my work on these calculations. Not all of you will want to read it, so I want to sum it up as concisely as possible. The long and the short of it is proc sets are OP, just as we suspected all along, and ZOS's fix of removing crit from them made nearly no difference in them. If I were to give rough estimates of the weapon and spell damage equivalents of the proc sets Viper, Selene, and Skoria, it would be at roughly 900, 1900, and 800 unbuffed, over their duration. Presently, I don't have the applicable ranges figured out yet, but know that these numbers go up (quite a lot) when using weaker attacks and go down (on a much smaller scale) as you crit higher and more frequently. I am working on an equation to predict the weapon damage equivalence of proc sets in a given scenario. The only things that I am really missing to verify these calculations are parses from in-game and someone to check my work and thought processes. If you still want to read the rest of this, well, enjoy.

---

Proc sets have been around for a long time, and for a long time we have been saying they were overpowered. Unfortunately, because of how they work, it is difficult to actually measure how strong they are. There are already innumerable interacting variables that affect damage and no one wants to go to the trouble to figure these things out quantitatively. I don't even want to do this, it's a mess and very time consuming and rather imprecise. I also had been acting under the assumption that someone had already gone to the trouble to do this and would soon post the results.

TLDR, this paragraph is somewhat off-topic and you can disregard it:
Well, today I was bored and thought about the last week. Before this week, the only spec I ever played was magicka and the only class I ever really touch is Templar. I got bored with that on Monday and decided to mess around with one of my many toons that I had already leveled to 50, my bosmer stamina dragonknight. I have been buying sets and trying out many combinations trying to get something that fit my playstyle, because of course the end goal is to 1vX (in this case, in Sotha Sil). It was fun, though very weak for many reasons, but I can tell when a set is or isn't performing to my needs. I wanted a very high damage build without crippling my defense nor sustain, like my magplar's. High, even unreasonable bar to set, I know. The best place to start was the meta, and, well... skipping the details I tried many sets that just couldn't meet my standards. The only thing I had not looked at were proc sets, because I absolutely refuse to use proc sets. It is my opinion that they are so overtuned and anyone using them is crutching and not worth serious consideration as a great player. Without real evidence, however, my feelings on that aren't really worth a damn (wont change how I feel though). Tired of waiting for the analysis to be dropped into my hands from someone like Asayre, I got off my lazy arse (figuratively, not literally. I was seated the entire time) and got to work.

My first stop, Sorcerer Arithmagic, led me to the UESP build editor which I trust to be very reliable and accurate, if not incredibly frustrating to navigate (don't hit the backspace button or you may lose everything). Now, earlier today I did a few calculations (and the results will shock you!) but let's take a new example, fudge it a little bit, and hope it still works out. If not, I'll type up my results of previous examples in order to get a point across or something. I have not run these numbers yet, they will be new to me as I type them. Let's consider an Imperial Stamina Dragonknight. This sDK will be a total scrub in PvP and he wants to make a build that can deal the most damage possible while spamming Wrecking Blow. He decides on Clever Alchemist and Fury because of the sheer amount of weapon damage available in them.

Starting from scratch, a bare naked player has 7958 Stamina and Magicka. He puts 64 attributes into Stamina, which add 111 each, so adding 7104 Stamina onto his base for 15062 Stamina points. Somehow, this scrub has got his hands on perfect gold versions of Fury and Clever Alchemist. In fact, he is a developer and hacked the Live game to have the Fury jewelry have robust traits in legendary (gold) quality. These jewels have 870 Stamina each, which adds 2610 Stamina to his base for a running total of 17672 Stamina. Because this player has a modicum of sense, all of his traits are Impenetrable, because he wants to live through the crits he sustains. All 7 pieces of gear have gold stamina enchants on them, adding 868 on the three large pieces and 351 on the four small pieces, which amounts to 4008 Stamina bringing the total to 21680 Stamina. Fury has a single stamina set bonus of 967 at full golded quality, topping his base Stamina off at 22647.

The Scrub-onknight now equips his sharpened gold battle-axe of fury, because as you can imagine he never got high enough in a campaign to get a nirnhoned one, and he wasn't thinking about the benefits of a sword or a mace. O well, his loss. His weapon damage is 1571. That's okay though, because he has used a dev command again to give himself 100% uptime on clever alchemist, and he found a bug with caltrops that let's him take a crit from that every second forever, wherever. He neglected to share or fix this bug. Clever Alchemist contributes 790 weapon damage and Fury adds another 879. He also has 3 golded weapon damage enchants on his jewelry which combine to add 522 weapon damage, bringing his base weapon damage to a total of 3762. Crazy!

Now, this reinforced mantis shrimp is wearing 5 medium armor and has excellent uptime on both forms of brutality because he is a DK, so we multiply 3762 by 1.37 as see him walking around with a whopping 5154 weapon damage. His stamina, because he is wearing 1 piece of heavy, is boosted 14% through being an imperial and using undaunted. This brings it to 25817. It's somewhat low, but that's because his food ran out! He chugs 5 Dubious Camoran Throne, just to be safe, for 3192 Stamina, bringing the final value of it to 29010. Did I round? Slightly. Why? Get over it. If he can cheat, I can round to the ones place.

According to UESP, the tooltip for Uppercut is equal to 0.14935S + 1.56805W + 2.14396, with S being Stamina and W being Weapon Damage. Let plug these numbers in and... we get 12,416. I am doing this with you and, well, I actually expected higher. But remember, this is no CP. OK. Moving on!

So the other devs found and fixed his precious caltrops bug, so the Dumpsterknight decided to switch out Fury for a more reliable set. Maybe one he can gank with. He has been silently reading the forums and noticed how everyone is calling proc sets OP, so he thinks he must have it for himself. He gets a full set of viper, golded, perfect, with jewelry. He looks at it. He sees he is trading his hacked Fury for a legit Viper and the only real difference in damage output is the trade of 879 weapon damage for 6400 poison damage every 4 seconds. So, his weapon damage drops from 5154 to 3950. Quite a lot! Is it worth the viper proc? Even I don't know yet! Let's find out...

Using the equation above, the tooltip comes out to... 10528.

So lets put him into a fight now, shall we? You have found yourself in the unfortunate position of having 0 resistances. This could be a bug, or it could be corrosive armor. Doesn't matter. In four seconds of you derping off in a load screen, he hits you with a WB spam. If he has Fury on, you would have taken 37249 damage in three hits, or 49666 in four hits. If he has Viper on, you would take 6400 damage on the first hit and then 31585 damage from the 3 hits of uppercut, for a total of 37985 damage. In four hits, you'd take 48514 damage.

Wow. I did not honestly expect that. By the fifth hit, viper will be in the lead again. This was also assuming robust fury jewels, which do not exist. I think this speaks for itself.

Note that these numbers have assumed no crits and battle spirit has not affected them yet. This is not to say you must be dead from this damage, or whatever. The key tidbit of information to get out of this is the effective weapon damage that proc sets, in this case specifically viper, have. Perhaps this will convince people (devs) that they are absurdly high? Ah, whatever. As I said above, if my assumptions and calculations are correct, this should speak loudly and clearly. If you notice any errors or improper reasoning, please bring it up. I made a number of them in the first few tries of this earlier today; after all, I'm not Asayre.

NightbladeMechanics

I will read this after PvP. I know you've been doing some deep digging into the weapon damage equivalent of Viper tonight, but I still think you should do those 3 dps parses and compare.

1) Hunding's or Viper 4pc, no other gear.

2) Hunding's 5pc, no other gear.

3) Viper 5pc, no other gear.

Their 2-4pc bonuses are the same, so these three parses would isolate contributions from the 5pc bonuses alone. And since Hunding's is a pure weapon damage 5pc, you can draw a line between the dps of the 4pc (0 weapon damage), the dps of Hunding's (300 weapon damage), and then plot the dps of Viper on that line to isolate its weapon damage equivalent.

WhiteMage

Well I'm not about to grind and gold out things I will never use, so I can't do a parse (totally can't stam deeps in pve properly, and this is a pvp thing) so I'll do the next best thing! I'll apply the same as what I did above for your situations.

Here are the assumptions:
5 pieces of medium CP 160 armor
Using a gold 2h that a part of no set, also CP 160
The armor will include chest, legs, gloves, belt, and shoes with perfect stamina glyphs on them
The attacks will not crit
No CP is used, but all attribute points are
The character is an imperial sDK
Tri-food is used
Wrecking blow will be spammed

The base stamina is 7958 + 7104 + 2739 (5 stam enchants, 2 large, 3 small) + 967 (1 stamina set bonus) = 18768
The base weapon damage is 1571.

Stamina is increased 12% due to racial and undaunted: 21020
Stamina is increased further to a final value of 25125 with food (+4105).
Weapon damage is increased 37% due to major and minor brutality and the medium armor buff: 2152

If we plug these numbers as is into the equation for uppercut (0.14935S + 1.56805W + 2.14396), we get that one uppercut deals 7129 damage.

Applying the Hunding's Rage set bonus of 300 weapon damage, that increases to 7773 damage. Let's throw three and four of those out there against something with zero resistance and HR deals 23320 and 31093 damage, respectively.

Take the first one again, use it three and four times plus one full viper proc deals 27787 and 34916 damage, respectively. Without the viper proc, these numbers would be 21387 and 28516.

Using ever so slightly more complicated math which for some reason I neglected to apply to the original post, on the three hit sample the effective weapon damage increase due to viper is 1360 (993 base weapon damage before buffs) while on the four hit sample the weapon damage is increased 1022, or 745 base weapon damage before buffs.

WhiteMage

Going back to the example in the OP, we can calculate the effective weapon damage increase of viper when used in conjunction with Clever Alchemist. The fully buffed stamina and weapon damage values are 29010 and 3950, respectively. A single uppercut does 10528 like this. Looking at this with one, two, three, and four hits under a single viper proc, we get 16928 damage (1 hit), 27457 damage (2 hits), 37985 damage (3 hits), and 48514 damage (4 hits). These are consistent.

Now let's take each number and divide by the number of hits for each. They are, respectively: 16928, 13728, 12662, and 12128.
Then we subtract each number by 4334, which is the contribution from stamina plus that constant (29010 * 0.14935 + 2.14396): 12594, 9394, 8328, and 7794.
Lastly, we divide by the weapon damage coefficient of 1.56805, yielding: 8032, 5992, 5312, and 4971. These numbers are the effective weapon damages we are using in these situations with viper. Let's subtract what we already knew was there, the 3950, and get: 4082, 2042, 1362, 1021.

Once again, I am surprised. This time by how it seems to scale exactly the same way as the other example...

Lastly, divide each number by 1.37 to find the effective base weapon damage increase with each. Remember that Hunding's is only 300. They are: 2979, 1490, 994, 745.

So it seems that Viper is equivalent to 745 weapon damage, If you are spamming uppercut and if you can manage to get 4 attacks in before the next proc. Most of the time, I can only fit 3. Please note that there are no DoTs running on the victim and no animation canceling, which I believe would lower the effective weapon damage if they were present, but not nearly to the point of equivalence with hunding's. I don't have numbers to back that up, at the moment, of course.

sly007

I read it all. So proc sets are over performing.

WhiteMage

More thoughts. If all else (including stamina) remains equal, it appears that the weapon damage increase from viper (or any proc set?) can be divided by the cooldown (for the lowest possible contribution) or some other specified time and then divided further by the weapon damage coefficient to get the effective weapon damage increase. For example, if instead of uppercut we used flurry, whose weapon damage coefficient is 0.13494 (flurry deals 8 times the tooltip damage so lets just multiply that by 8 to get 1.07952), we take 6400 damage from viper, divide that by 1, 2, 3, and 4 to get 6400, 3200, 2133, and 1600, then divide each of those by 1.07952 to get an effective buffed weapon damage increase of 5928, 2964, 1976, and 1482. This makes sense for it to be higher than uppercut because flurry gets a lesser contribution to its damage from weapon damage (confusing terms, right? ZOS named them, not me.). This suggests that viper is even more effective when using weaker skills, which lends itself to an artificially increased floor. Did you ever feel that bad players suddenly became better more dangerous when they put on a proc set? There may be some credence to that.

Kas

good approach, but I dislike the assumptions:

1) crits DO matter, a lot. Even against impen.
2) so do light attack weaves (and bashes) and so do dots.
3) the higher weapon damage also increases healing from vigor and rally
not important) viper is poison damage. makes it slightly worse against Argonians but can proc the poisoned status effect with a very small chance

but then, spamming attacks makes is an assumption very favorable towards viper. under pressure and against moving targets, you'll get in, on average, so far less attacks than 3 or 4 between viper cooldowns.

imho this makes it very hard to compare . personally I arrived at the following conclusion:

1) viper is a great set if you run a dual wield frontbar and can complete it there, combining it with undaunted monster set and another 5-piece.
2) if you run a s&b bar, this is possible too. but often s&b is the defensive bar and alchemist can fulfil that role better
3) there are other sets I personally prefer that do not contribute towards raw damage output. e.g., homestead bonepirate was insane
bonus) viper is bad for large-scale group pvp - but so are all stamina builds that aren't speedbuff spamming support, tbh

WhiteMage

Excellent, some discussion. First off, you are right that weaves and bashes and DoTs do matter and they will lower the effective weapon damage increase of viper, which is in the more-balanced as opposed to less-balanced direction. If you have some way to quantify the benefits to a PvP build of damage and healing, please share it, because otherwise we have no choice but to ignore it. Proc sets wouldn't be the only things to suffer a penalty in the such a metric, however. A very many other popular and very effective sets like War Maiden, Sun, and Automaton would be rated lower, even lower than proc sets like viper because they work on more than a single damage type. In that sense, they don't follow the postulated trend of the more niche being more powerful and vice versa.

Spamming the same attack does favor proc sets more, as acknowledged above, but this assumption was made both for the sake of simplicity and generality, on top of being a not-uncommon occurrence in-game. I do disagree, however, when you say that being under pressure and not getting the opportunity to land attacks on cool down harms my case. The fewer attacks you are able to get in, the more powerful proc sets are so long as you can proc them near when their cool down lapses, but even if you can't your dps drops faster than the effective weapon damage contribution. Landing one attack every second gives the full benefit of doubt to stat increasing sets, similar to how cost reduction become more effective as you increase the frequency of skill usage. In PvP where you very likely do not get to land 4 consecutive and uninterrupted attacks in four seconds viper is favored.

As for your conclusion (the third one), in this thread I posted about the two extreme situations, a full damage set and no sets with viper. The result was almost identical. The other situation I mentioned in the OP but did not post was in fact with bone pirate, and the results were the same (even by the squeeze theorem we could guess that would be true). Viper does not appear to become more or less effective with more or less weapon damage. It does not depend on it. As for your note about effectiveness in a large group environment, that is entirely beside the point because much to the same degree all similar specs, which are designed for 1vX and small-scale, are less effective than group oriented builds in group situations, and the reverse holds painfully true as well. A group spec without a group is a fish out of water, and can do little but flail uselessly until it dies a quick but painful death.

Lastly, I PvP in no CP, where people don't get an extra free 12% crit chance and I find crits to be a non-factor because I wear full impen. Including them in such short parses as the duration of most fights in small-scale is not so simple as the calculation you would make in PvE to find the effective damage and is even further complicated when you use more than one skill. Nevertheless, if you can provide me a spec and PvP "rotation" and it's duration, I will try to figure out its effective viper-substituted weapon damage increase (or decrease, though I very much doubt that would happen). I will even include crits, so be as detailed as possible, please.


In PvP, I really don't like crit because you can't rely on it. It's just... useless imo.

Brutusmax1mus

The potential of viper is dependent on the user. The burst it provides is outstanding in PvP and mediocre at best in pve.

Viper is also cheesy, as is anything that doesn't require player commands to perform. Procs and threshold occurrences like mages fury or implosion are bad design and one of the reasons the PvP on this game will never be an esport.

Waffennacht

Yes

Lexxypwns

WhiteMage wrote: »

Excellent, some discussion. First off, you are right that weaves and bashes and DoTs do matter and they will lower the effective weapon damage increase of viper, which is in the more-balanced as opposed to less-balanced direction. If you have some way to quantify the benefits to a PvP build of damage and healing, please share it, because otherwise we have no choice but to ignore it. Proc sets wouldn't be the only things to suffer a penalty in the such a metric, however. A very many other popular and very effective sets like War Maiden, Sun, and Automaton would be rated lower, even lower than proc sets like viper because they work on more than a single damage type. In that sense, they don't follow the postulated trend of the more niche being more powerful and vice versa.

Spamming the same attack does favor proc sets more, as acknowledged above, but this assumption was made both for the sake of simplicity and generality, on top of being a not-uncommon occurrence in-game. I do disagree, however, when you say that being under pressure and not getting the opportunity to land attacks on cool down harms my case. The fewer attacks you are able to get in, the more powerful proc sets are so long as you can proc them near when their cool down lapses, but even if you can't your dps drops faster than the effective weapon damage contribution. Landing one attack every second give the full benefit of doubt to stat increasing sets, similar to how cost reduction become more effective as you increase the frequency of skill usage. In PvP where you very likely do not get to land 4 consecutive and uninterrupted attacks in four seconds viper is favored.

As for your conclusion (the third one), in this thread I posted about the two extreme situations, a full damage set and no sets with viper. The result was almost identical. The other situation I mentioned in the OP but did not post was in fact with bone pirate, and the results were the same (even by the squeeze theorem we could guess that would be true). Viper does not appear to become more or less effective with more or less weapon damage. It does not depend on it. As for your note about effectiveness in a large group environment, that is entirely beside the point because much to the same degree all similar specs, which are designed for 1vX and small-scale, are less effective than group oriented builds in group situations, and the reverse holds painfully true as well. A group spec without a group is a fish out of water, and can do little but flail uselessly until it dies a quick but painful death.

Lastly, I PvP in no CP, where people don't get an extra free 12% crit chance and I find crits to be a non-factor because I wear full impen. Including them in such short parses as the duration of most fights in small-scale is not so simple as the calculation you would make in PvE to find the effective damage and is even further complicated when you use more than one skill. Nevertheless, if you can provide me a spec and PvP "rotation" and it's duration, I will try to figure out its effective viper-substituted weapon damage increase (or decrease, though I very much doubt that would happen). I will even include crits, so be as detailed as possible, please.


In PvP, I really don't like crit because you can't rely on it. It's just... useless imo.

I'm sorry, I was very interested in what you had to say until you said crit was useless. For morrowind, the best 2 ways to build monster burst now that we're forced into dropping damage for sustain are; Either you use proc sets to create large amounts of concurrent damage or you use high crit + shadow.

GreenSoup2HoT

Lexxypwns wrote: »

WhiteMage wrote: »

Excellent, some discussion. First off, you are right that weaves and bashes and DoTs do matter and they will lower the effective weapon damage increase of viper, which is in the more-balanced as opposed to less-balanced direction. If you have some way to quantify the benefits to a PvP build of damage and healing, please share it, because otherwise we have no choice but to ignore it. Proc sets wouldn't be the only things to suffer a penalty in the such a metric, however. A very many other popular and very effective sets like War Maiden, Sun, and Automaton would be rated lower, even lower than proc sets like viper because they work on more than a single damage type. In that sense, they don't follow the postulated trend of the more niche being more powerful and vice versa.

Spamming the same attack does favor proc sets more, as acknowledged above, but this assumption was made both for the sake of simplicity and generality, on top of being a not-uncommon occurrence in-game. I do disagree, however, when you say that being under pressure and not getting the opportunity to land attacks on cool down harms my case. The fewer attacks you are able to get in, the more powerful proc sets are so long as you can proc them near when their cool down lapses, but even if you can't your dps drops faster than the effective weapon damage contribution. Landing one attack every second give the full benefit of doubt to stat increasing sets, similar to how cost reduction become more effective as you increase the frequency of skill usage. In PvP where you very likely do not get to land 4 consecutive and uninterrupted attacks in four seconds viper is favored.

As for your conclusion (the third one), in this thread I posted about the two extreme situations, a full damage set and no sets with viper. The result was almost identical. The other situation I mentioned in the OP but did not post was in fact with bone pirate, and the results were the same (even by the squeeze theorem we could guess that would be true). Viper does not appear to become more or less effective with more or less weapon damage. It does not depend on it. As for your note about effectiveness in a large group environment, that is entirely beside the point because much to the same degree all similar specs, which are designed for 1vX and small-scale, are less effective than group oriented builds in group situations, and the reverse holds painfully true as well. A group spec without a group is a fish out of water, and can do little but flail uselessly until it dies a quick but painful death.

Lastly, I PvP in no CP, where people don't get an extra free 12% crit chance and I find crits to be a non-factor because I wear full impen. Including them in such short parses as the duration of most fights in small-scale is not so simple as the calculation you would make in PvE to find the effective damage and is even further complicated when you use more than one skill. Nevertheless, if you can provide me a spec and PvP "rotation" and it's duration, I will try to figure out its effective viper-substituted weapon damage increase (or decrease, though I very much doubt that would happen). I will even include crits, so be as detailed as possible, please.


In PvP, I really don't like crit because you can't rely on it. It's just... useless imo.

I'm sorry, I was very interested in what you had to say until you said crit was useless. For morrowind, the best 2 ways to build monster burst now that we're forced into dropping damage for sustain are; Either you use proc sets to create large amounts of concurrent damage or you use high crit + shadow.

Agree'd. Players vastly underestimate crit damage. Especially those who use shields. Even though you cannot crit shields, once you break through them crit is very good.

NightbladeMechanics

I've been on a tirade about crit damage being the highest scaling damage stat for what, 3 patches now? It's sooooo important always.

And any color quality of viper and hundings will work as long as the glyphs and traits and colors are all the same across the two sets. We need to find someone to do those parses.

WhiteMage

Don't discount my reasonable, mathematically supported conclusions just because I have opinions that are largely unsubstantiated lol. If anyone has a way to quantify it, please share. Regardless, though, I will probably never build around it. I do, however, recognize that this is has a significant enough of an impact to warrant defending against. I use 7 pieces of impenetrable gear and can count the times I thought "Wow, that's a high crit in my death recap. No wonder I died," on one hand. My aversion to it is nuanced.

ANYWAY

Let's try something neat. I'm going to take my personal magplar build, which consists of Rattlecage, Kagrenac's Hope, and Grothdarr and swap Kagrenac's Hope out for Viper. I will still be using all magicka skills and I'll even change the viper pieces (5 body slots) into stamina enchants, instead of the triglyphs and magicka glyphs I actually use. The whole thing is courtesy of UESP's build editor. Once again, I don't know how this will turn out as I am typing this, but I have a hunch that it will favor viper, a stamina set, while playing a magicka specced toon.

So. On Live my Magicka is 30388. This is in a no CP environment, just for clarification. My weapon damage is, fully buffed, 2891(2775 pseudo-unbuffed). My sweeps tooltip is 906, my toppling tooltip is 5007, my RD tooltip is 10940, my DBoS tooltip is 9338 with a DoT of 10476, and in case youre interested, my HtD tooltip is 9663. In a short fight against a potato, my target will be dead in six seconds or less following a combo something like Backlash>Toppling>Jabs>Jabs>DBoS>RD (Backlash burst). Let's take the Toppling>Jabs>Jabs>DBoS part into consideration only because it is predictable and UESP agrees with it. These numbers assume burning light procs exactly once per jab, and there are no crits and Battle Spirit is ignored for the purposes of this. I deal 38432 damage during these four attacks (Burning Light tooltip is 3340 and the DBoS DoT is not allowed to tick). If I want to light attack weave during this, then the total damage rises to 42164, which assumes 3 light attacks with the damage they do on Live, 1244.

Ok, so now I trade Kags for Viper. This is an exchange of roughly 222 base spell damage and one magicka set bonus for one stamina set bonus and a guaranteed viper proc. According to UESP, my stats drop to 28787 Magicka and 2620 Spell Damage, and the resulting tooltips for these four attacks are: 840 Jabs tooltip, 3096 Burning Light tooltip, 4641 Toppling tooltip, and 8655 DBoS direct damage tooltip. Math it up and throw in one viper proc and we have 42016 damage. If we throw in 3 light attack weaves at the Live value (if anything they would go up when switching to viper) my Magplar would have dealt 45748 damage.

This suggests that Viper would be worth more than 222 spell damage for my magplar. Will I make the switch? Probably not because I would trade a hefty amount cost reduction and the like by going from light to medium. It looks like it would be a damage increase though. What does this say to me? It says proc sets are king.


EDIT: Whoops! Silly me, in everything so far I have used 6400 as the damage from viper. According to UESP, its actually 6720 when perfect. So... the margin on everything I've done so far shifts slightly to favor viper more.

WhiteMage

This time, instead of discussing the weapon damage contributions of viper, I am going to try to compare the two monster sets Selene and Kena. Just as a refresher for everyone, a perfect Selene set provides one set bonus of stamina and a 15% chance to proc 12000 physical damage on a melee attack every four seconds. On the other hand, a perfect Kena set gives a set bonus of 129 weapon and spell damage and 516 more weapon and spell damage for 6 seconds, but at the cost of a 20% skill cost increase. It is procced by 2 consecutive light attacks. This will be on a Redguard Stamina Sorcerer build using 2h/SnB of viper and [redacted]. This build is, on paper, very viable in PvP. Everything is gold ([redacted]) with weapon damage and stamina enchants. It is 5 medium and 2 [redacted], so the undaunted bonus is 4%. Trifood is used. In this comparison, Kena is assumed to have a 100% uptime with no need to proc it. Selene is assumed to proc every six seconds. This stamina sorcerer will, of course, have Hurricane running and be charging into the fight with a Crit Rush>DS>DBoS>Executioner. The thing about executes is, first, I don't know the scaling, and second, differences in damage dealt leave the target with different health percentages... so this is not something I will include in this theoretical scenario. We can assume, though, that the target potato is dead after the execute. We've all seen this combo before if not been a victim to it ourselves. This Sorcerer has two Storm Calling abilities and one Fighter's Guild ability slotted on this bar, for +7% additional weapon damage. Major Brutality, but not Minor Brutality, is active.

According to UESP, with a procced Kena this sorcerer has 3805 Weapon Damage and 30717 Stamina. With Selene, he instead has 2909 Weapon Damage and 31819 Stamina. First off, no crits. The tooltips for the relevant skills are:

Dizzying Swing: 0.15367S + 1.61381W + 2.72219 <--- Notice the difference between previous posts? That's me not being thorough. What that translates to is the above posts using the unmorphed Uppercut instead of the morphed version.
Critical Rush: 0.07156S + 0.75102W + 4.98990
Hurricane: 0.01059S + 0.11120W + 2.34615
DBoS direct damage hit: 0.15382S + 1.6135W - 0.11867

I'll go ahead and do the math... for Kena:
DS: 10863
CR: 5060
Hrcn: 750
DBoS direct damage: 10864

And now for Selene:
DS: 9587
CR: 4466
Hrcn: 663
DBoS direct damage: 9586

Now let's put them all together. Hurricane deals up to 150% more damage, which means its tooltip gets multiplied by 2.5. We will assume all three ticks of hurricane deal that maximum damage. In both cases, viper will proc once and selene will proc only once also. The damage that this set deals with Kena is: 39132 (viper proc is 6720, I remembered this time). With Selene, 47331 damage is dealt. For the sake of argument, if we increase the direct damage hits by, say, 25%... the gap becomes larger.

Kena falls so far short of Selene in this case. How much weapon damage does it provide? Kena grants 645 base weapon damage to the user which is increased by 39% by this sorcerer to 896 weapon damage. Kena also has a detriment of 20% cost increase, perhaps to limit how powerful it is. No proc sets have a downside or tradeoff for their power.

Ok, this sorcerer has only 23.8% crit, but for the sake of argument, every single one of his attacks will crit. In an attempt to balance proc sets, ZOS removed their ability to crit. Let's see how much this helps. With Kena, 55338 tooltip points of damage will be dealt, but Selene... reaches 61637. The damage gap was lowered by roughly 2000 when EVERYTHING crit with a modifier of 1.5.

I don't have the numbers to back this up Thanks to my trusty spreadsheet skillz, it appears that Selene amounts to roughly 2600 (if nothing crits) points of weapon damage (~1900 unbuffed), at least in a short window. If they are indeed balanced anywhere, it would be in PvE where the numbers work out to be roughly equivalent as the encounter stretches towards infinity. The may very well have balanced dps, but their burst potential has no rival.

EDIT: Per request, I removed some information on the secondary set. Of note, it contributes only 1 stamina set bonus to damage.

NightbladeMechanics

Jesus, that's op too.

WhiteMage

WhiteMage wrote: »

So lets put him into a fight now, shall we? You have found yourself in the unfortunate position of having 0 resistances. This could be a bug, or it could be corrosive armor. Doesn't matter. In four seconds of you derping off in a load screen, he hits you with a WB spam. If he has Fury on, you would have taken 37249 damage in three hits, or 49666 in four hits. If he has Viper on, you would take 6400 damage on the first hit and then 31585 damage from the 3 hits of uppercut, for a total of 37985 damage. In four hits, you'd take 48514 damage.

Let's go ahead and assume everything crits, which it won't, of course, but let's see what happens. If he is using Fury, three crits at a crit modifier of 1.5 would deal 55873, four hits would deal 74499 damage. With viper on instead, and adjusting for the proper value of viper, three hits would amount to 54097 damage and four hits would deal a total of 69891. I'm not sure why I didn't just put this in the original post (it's because I didn't think of it), but here are the numbers you get with the two most damaging sets critting with every single attack. Viper will get closer on the next hit but it will not outpace the damage again.

This may suggest that viper is overtuned by about 50%, so drop it down a third to 4480 and it may be balanced. Maybe it could even be allowed to crit at that level. MAYBE. Do NOT think that this is the right solution for viper, there are no numbers to back that up here. This paragraph is just speculation.

Lexxypwns

WhiteMage wrote: »

Don't discount my reasonable, mathematically supported conclusions just because I have opinions that are largely unsubstantiated lol. If anyone has a way to quantify it, please share. Regardless, though, I will probably never build around it. I do, however, recognize that this is has a significant enough of an impact to warrant defending against. I use 7 pieces of impenetrable gear and can count the times I thought "Wow, that's a high crit in my death recap. No wonder I died," on one hand. My aversion to it is nuanced.

ANYWAY

Let's try something neat. I'm going to take my personal magplar build, which consists of Rattlecage, Kagrenac's Hope, and Grothdarr and swap Kagrenac's Hope out for Viper. I will still be using all magicka skills and I'll even change the viper pieces (5 body slots) into stamina enchants, instead of the triglyphs and magicka glyphs I actually use. The whole thing is courtesy of UESP's build editor. Once again, I don't know how this will turn out as I am typing this, but I have a hunch that it will favor viper, a stamina set, while playing a magicka specced toon.

So. On Live my Magicka is 30388. This is in a no CP environment, just for clarification. My weapon damage is, fully buffed, 2891(2775 pseudo-unbuffed). My sweeps tooltip is 906, my toppling tooltip is 5007, my RD tooltip is 10940, my DBoS tooltip is 9338 with a DoT of 10476, and in case youre interested, my HtD tooltip is 9663. In a short fight against a potato, my target will be dead in six seconds or less following a combo something like Backlash>Toppling>Jabs>Jabs>DBoS>RD (Backlash burst). Let's take the Toppling>Jabs>Jabs>DBoS part into consideration only because it is predictable and UESP agrees with it. These numbers assume burning light procs exactly once per jab, and there are no crits and Battle Spirit is ignored for the purposes of this. I deal 38432 damage during these four attacks (Burning Light tooltip is 3340 and the DBoS DoT is not allowed to tick). If I want to light attack weave during this, then the total damage rises to 42164, which assumes 3 light attacks with the damage they do on Live, 1244.

Ok, so now I trade Kags for Viper. This is an exchange of roughly 222 base spell damage and one magicka set bonus for one stamina set bonus and a guaranteed viper proc. According to UESP, my stats drop to 28787 Magicka and 2620 Spell Damage, and the resulting tooltips for these four attacks are: 840 Jabs tooltip, 3096 Burning Light tooltip, 4641 Toppling tooltip, and 8655 DBoS direct damage tooltip. Math it up and throw in one viper proc and we have 42016 damage. If we throw in 3 light attack weaves at the Live value (if anything they would go up when switching to viper) my Magplar would have dealt 45748 damage.

This suggests that Viper would be worth more than 222 spell damage for my magplar. Will I make the switch? Probably not because I would trade a hefty amount cost reduction and the like by going from light to medium. It looks like it would be a damage increase though. What does this say to me? It says proc sets are king.


EDIT: Whoops! Silly me, in everything so far I have used 6400 as the damage from viper. According to UESP, its actually 6720 when perfect. So... the margin on everything I've done so far shifts slightly to favor viper more.

Viper is halved by battle spirit.

Regardless of the straw man you create against crit, it is a very reliable way to get damage. Take a magplar kill combo. You've got a reflective light tick, charge, jabs, burning light, DBoS purifying light burst, beam. That's ~10 sources of damage in less than 2 seconds. If half of those can reliably do 80% more damage than your burst becomes significantly more reliable.

Proc sets work in a similar way, adding to your burst. This has great synergy with using crit to increase burst, which is why proc sets can no longer crit. Even still, being able to reliably make half your burst hit 75%+ harder AND stacking a 3k viper and 5k Selene is where the real burst comes from

NightbladeMechanics

I am going to add that even if Viper contributed only an amount of damage in long parses to equal Hunding's 300 weapon damage, it would still be the stronger pick in PvP because 1) it frontloads that damage into the first global cooldown of its 4 second cooldown, and 2) its damage doesn't cost resources.

I'll also add that comparing procs to weapon damage or max stam bonuses in combos or parses containing fewer numbers of abilities will weigh in favor of the procs. If you add layered dots or more abilities into the combo, weapon and spell damage will have more sources to amplify, potentially closing the gap. This is why NMG and Hunding's are preferred to Viper in PvE.

But realistic PvP combos contain fewer instances of damage than dps parses. All hail the proc set master race!

NightbladeMechanics

@WhiteMage you're a poet by the way. I need more story times.

Mage and @Lexxypwns , we can't isolate a weapon damage equivalence for Viper proc without first defining the parse.

Because viper's damage is flat, the more abilities and dots you layer together, the more Hunding's (or any flat stat bonus) will close its gap behind Viper. In short, Viper has no weapon damage equivalent. It has a flat dps equivalent, normalized for its cooldown. Derp.

That makes Mage's thought experiments very good, actually. He's defining different parses done by different classes and comparing the total damage contribution of Viper to that of the next best (or at least another common) set. There are a ton of variables at play, though. This makes these scenarios just that -- thought experiments, not scientific measurements. He acknowledged this at the start and has done a good job of mentioning the loose ends and how they affect his comparisons. .

@WhiteMage can you draw up a scenario where the same character in the OP plays his equally questionably geared and arguably more op stamina scrubblade? I'd like to see a comparison between a full 2h heavy attack > Incap > Surprise Attack > Executioner while wearing Hunding's and Viper. Bonus points if you calculate for Follow Up.

Lexxypwns

NightbladeMechanics wrote: »

@WhiteMage you're a poet by the way. I need more story times.

Mage and @Lexxypwns , we can't isolate a weapon damage equivalence for Viper proc without first defining the parse.

Because viper's damage is flat, the more abilities and dots you layer together, the more Hunding's (or any flat stat bonus) will close its gap behind Viper. In short, Viper has no weapon damage equivalent. It has a flat dps equivalent, normalized for its cooldown. Derp.

That makes Mage's thought experiments very good, actually. He's defining different parses done by different classes and comparing the total damage contribution of Viper to that of the next best (or at least another common) set. There are a ton of variables at play, though. This makes these scenarios just that -- thought experiments, not scientific measurements. He acknowledged this at the start and has done a good job of mentioning the loose ends and how they affect his comparisons. .

@WhiteMage can you draw up a scenario where the same character in the OP plays his equally questionably geared and arguably more op stamina scrubblade? I'd like to see a comparison between a full 2h heavy attack > Incap > Surprise Attack > Executioner while wearing Hunding's and Viper. Bonus points if you calculate for Follow Up.

I worry that discounting crit exaggerates proc set strength since they already exist without crit

WhiteMage

@NightbladeMechanics Yes I can, however I can't currently include an execute that scales with lost health without knowing for certain the function. Instead of Executioner, I'll use Killer's Blade because it either scales to the full damage or it doesn't scale at all. This differently "questionably" geared Stamina Nightblade will be an Imperial wearing the bugged/hacked perma-procced Fury set with robust jewels and either Viper or Hunding's Rage (Fury contributes more to damage than Clever Alchemist, at an obvious cost which we are circumventing). UESP doesn't allow for these jewels, of course, so we'll do this the hard way, by hand.

We start with a base Stamina of 7958 and add 64 Stamina attributes for 7104 more, yielding 15062. Perfect enchants add 4008 and 3 robust traits add a further 2610 to reach 21680 Stamina. Both Viper and Fury have a Stamina set bonus of 967 each, bringing us to 23614. This Nightbaddie was carried through enough pledges to have maxed out undaunted and takes full advantage of it by wearing a light Velidreth epaulet and a heavy Fury helm with his 5 medium Viper. The stamina is boosted 16% (factoring in the race as well) to 27392 Stamina. Tri-food tops it off at 31497 Stamina.

Using a legendary 2h, the base weapon damage starts at 1571. Three gold weapon damage enchants boost this 522 more to 2093. Fury adds 879 more weapon damage to reach 2972. The single light Velidreth shoulder piece moves it up to 3101. Now we increase it 37% due to Major Brutality and the 5 piece medium armor bonus to 4248. If Hunding's is equipped instead of Viper, the weapon damage would be 4658.

So his damage-relevant stats are 31497 Stamina and 4248 or 4658 Weapon Damage, in no CP, depending on the set. Now, this sNB is not new to this, so he applies Minor Berserk before the gank. Because his first attack is from stealth, the Master Assassin passive increases his weapon damage 10% further for the first hit, and it is guaranteed a critical hit, and a stun to boot. Hemorrage grants 10% further critical damage on that hit... jeez nightblades are starting to scare me. I'm going to nerf him in the sense that he is too stupid to use a mundus. He never picked one up after the last reset like a year ago.

This first hit is annoyingly complicated. Let's see... I think it would be accurate to say the heavy attack deals 13805 and 14942 damage for Viper and Hunding's, respectively. Viper would proc now. I will assume it doesn't use nor consume the Follow-Up buff. No more crits will occur. The rest of these are very complicated and I don't know exactly how they work properly, namely the 20% incoming damage increase the incapped victim receives. I will assume the target is in execute range and pick the most damaging option, though I have reason to believe executes don't scale the same way as last patch. These execute numbers are between 1500-2000 off from each other, though, so it is minor. My calculations show that the viper ganker would deal 62440 damage before any and all mitigations and the hunding's ganker would deal 58661 damage with these same constraints. If the scaling of executes has changed, the Hunding's build would see the slightly larger decrease in damage mentioned. More critical hits would close this gap of 4000 damage, but by how much, I don't know.

For my part these numbers carry a lesser degree of certainty than the other simulations I have done, but if I dotted all my i's and crossed all my t's appropriately, they will be very near the ballpark.



One last note to a previous poster... Yes, Viper is reduced by Battle Spirit, but so is all damage. Because of that, I choose to ignore it for these calculations because they don't really matter. You can divide all these damage numbers (every single one!) by 2 and then further by whatever additional mitigations you may experience to translate it to numbers you would actually sustain in game. The gaps get smaller but the proportions persist.

Lexxypwns wrote: »

I worry that discounting crit exaggerates proc set strength since they already exist without crit

"Discounting crit" does "exaggerate" the strength of proc sets, or rather it maximizes its relative performance. If everything crits, this is minimized. Both situations are very unlikely and the correct valuation lies somewhere in between. I have so far explored two situations where everything crits, in posts #15 and #17. Post #15 showcases Selene and even when everything crits at the base crit modifier Kena doesn't come close. This is partly due to the fact the viper is one of the sets used. On post #17 the example in the OP is used. It is nearly impossible to gain more weapon damage than in that example but even then viper is only just overtaken. A higher critical modifier would raise the gap even further, but in an already extraordinary circumstance getting more critical damage modifiers while still allowing most attacks to critically hit is unreasonable.

Brutusmax1mus

Why are you ignoring crit? That just makes viper seem stronger since youre comparing it to things that can crit yet you're not calculating that.

raasdal

Well, i am as big a fan as any, of the number crunching. But to get a ballpark estimate you can just do below;

Damage / Cooldown

6500 / 4 = 1625

You can do that because most spammables are around 1.0 in WD modifier. And you can fit one hit per second.

This does not however take into accout DoTs, Light Attack weaving or Crits.

With above, my guess would be that Viper is equivalent of about 1000WD in a short 4 sec burst window, with someone NOT using an ultimate.

Did i get it right? Have not had time to read all the posts for the result.

I used above method to determine the value of a Damage enchant vs Power enchant, when they released Poisons (and updated all enchants).

WhiteMage

I am primarily a magicka player, so I do have an inherent bias, but in an effort to get past that a bit, let's take a look at magicka's favorite proc set: Valkyn Skoria. With a cool down of 5 seconds, a little meteor is procced that deals 9000 damage and 4000 splash damage. For this build we shall go into battle with 2 perfect and procced sets of Burning Spell Weave and Scathing Mage. All jewelry is Arcane, all enchants are spell damage and magicka on a Breton. BSW (Burning Spell Weave) gives 654 Spell Damage and 967 Magicka. Scathing Mage gives 516 Spell Damage and 967 Magicka. Base magicka for this Breton is 23614 and spell damage, while dual wielding, is 3444.

Let's go crazy and have this Breton be a magicka Sorcerer with the skills Ice Comet, Inner Light, Bound Aegis, Power Surge, and Crystal Fragments slotted. Any arbitrary non-set medium and heavy pieces are on the head and shoulders. This brings the Magicka boost, will all relevant passives, to an amazing 33% and the Spell Damage boost to, with Major Sorcery, 27%. Factor those in and our Sorcerer has 31406 magicka before CP, which we will throw an extra 6048 magicka on top of to get 37454, thanks to green food. The spell damage is 4374.

This special little sorcerer has enabled the latest and greatest version of Cheat Engine and uses this to fire off Ice Comets on every global cool down. Let's see what happens (besides the incoming ban).

Ice Comet's tooltip is equal to, according to UESP, = 0.1983*M + 2.08151*Sp + 3.29096
This Ice Comet's initial attack will deal 16535 Cold damage. If we fire off four of them on a constantly moving target (they don't take the DoT damage), we deal 66139 damage before all mitigation, 82674 damage with five of them. 82674 is also the damage that would be dealt if exactly two of the Ice Comets crit at a 1.5 crit damage modifier, or CDM. Now replacing the arbitrary non-set pieces with Valkyn Skoria and having a proc go off, what do we get? 75139 and 91674 damage, respectively.
How much spell damage was Valkyn Skoria worth here?

It turns out that in four no-crit attacks, that scoria proc is worth 1081 buffed spell damage (851 unbuffed) and in five attacks or four with 2 standard crits the scoria proc is worth 864 buffed spell damage (680 unbuffed). This, of course, is an unusual situation, so let's look instead at something slightly more reasonable. This sorcerer will be spamming procced frags instead of Ice Comets, with the same bar set up.

A Crystal Fragment's tooltip is = 0.15369*M + 1.61382*Sp + 5.56173
For this sorcerer, the frag hits for 12820 damage before the skill's proc, 14102 after. (Wow!) So, four hits is 56411 damage and five hits or four hits and 2 crits at standard CDM is 70514. Throw in a Valkyn Skoria proc (disregarding HOW it procced for now) and we reach 65411 damage and 79514 damage, respectively. The spell damage equivalent that would have produced the same damage without Valkyn Skoria would be 1267 buffed (998 unbuffed) for four no-crit hits and with 5 hits or four hits with two crits (standard CDM), 1014 spell damage buffed (798 unbuffed).

WhiteMage

@raasdal

So I started to look into a little bit of the algebra of how these weapon/spell damage equivalencies work and it appears your approximations are very much in the ball park. I make no guarantees about my accuracy, but I believe that this equation can be used to determine the weapon damage equivalent of a proc set in a (finite) situation involving crit:

(ceiling(n * GCD / PCD) * P * (1 + PTM)) / (Z * (1 + ATM) * (1 + x * CDM))

Where:
n = number of attacks
GCD = average time between attacks that skills are not used (global cool down, factor in positioning issues etc.) I like to use 1.3
PCD = proc cool down, or average time between procs
P = base damage of given proc
PTM = proc tooltip modifier (minor berserk?)
Z = sum of weapon damage coefficients of skills used, including dot ticks, and if applicable, multiply them by their individual after-tooltip bonuses
ATM = after-tooltip modifier that affect all skills (minor berserk)
x = critical hit chance
CDM = critical damage modifier (not multiplier): base CDM = 0.5

This is my approximation for the weapon damage equivalent when using a sequence of skills. It is not exhaustive. So far, animation canceling/weaving is not included. Note that for each skill, the weapon damage coefficient is between 0 and 2.5, most typically around or below 1.

Waffennacht

Did I just spend like 10 minutes reading something that all boils down to, "proc sets OP, especially viper"?

I'm like, "no @#$_&quot;

WhiteMage

Yes, you did. Sorry if I didn't make that clear right from the get-go. I'll edit the OP with a major TL;DR to save people the trouble once I finish up with this thing I'm doing right now. The goal of this is to hopefully convince some developers that proc sets are overpowered and need to be toned down. The bandaid of removing crit from procs does little to slow the bleeding on a sliced jugular.

Izaki

Waffennacht wrote: »

Did I just spend like 10 minutes reading something that all boils down to, "proc sets OP, especially viper"?

I'm like, "no @#$_&quot;

I know right, I thought the conclusion would be something more... intricate you know xD

@WhiteMage Nice job on the calcs, although I'm not sure whether Sorc Arithmagic is outdated or not, haven't checked that one in a long time since I know most of it by heart (lul) and been more interested in Asayre's more recent threads, but the Build Editor is updated for sure and the calcs seem decent.

At this point though, I think what you're missing is parses. That way you'll be able to get a conclusion for sure.

raasdal

WhiteMage wrote: »

@raasdal

So I started to look into a little bit of the algebra of how these weapon/spell damage equivalencies work and it appears your approximations are very much in the ball park. I make no guarantees about my accuracy, but I believe that this equation can be used to determine the weapon damage equivalent of a proc set in a (finite) situation involving crit:

(ceiling(n * GCD / PCD) * P * (1 + PTM)) / (Z * (1 + ATM) * (1 + x * CDM))

Where:
n = number of attacks
GCD = average time between attacks that skills are not used (global cool down, factor in positioning issues etc.) I like to use 1.3
PCD = proc cool down, or average time between procs
P = base damage of given proc
PTM = proc tooltip modifier (minor berserk?)
Z = sum of weapon damage coefficients of skills used, including dot ticks, and if applicable, multiply them by their individual after-tooltip bonuses
ATM = after-tooltip modifier that affect all skills (minor berserk)
x = critical hit chance
CDM = critical damage modifier (not multiplier): base CDM = 0.5

This is my approximation for the weapon damage equivalent when using a sequence of skills. It is not exhaustive. So far, animation canceling/weaving is not included. Note that for each skill, the weapon damage coefficient is between 0 and 2.5, most typically around or below 1.

@WhiteMage - That is just awesome work Well done!

Lord-Otto

Here's the thing, VIPER is OP. It's unavoidable. Infernal Guardian is sooo easy to avoid. Even Velidreth and Selene can be avoided often enough. Skoria has really low proc chance and terrible proc conditions. It's just that silly 500% buff from SotH to Viper that made it so cancerous, since no counterplay.

Thanks for calculating, Mage!