Nighthold DPS breakdown: redefining the debate on class balance (with data)

[ufta]

Even if I reach my simulated 720k DPS for the Krosus mythic fight (which I did or one of my kills with good RNG a few weeks back) it does not even put me at the 70th percentile. Why? Because I have only one of the 3 best DPS legendaries.

You're building the entire basis of your argument on a faulty foundation. Sim dps has exactly ZERO to do with the discussion at hand. Your statement completely disregards the bell curve in SimC's results. That 720k doesn't mean that it's the max possible DPS for your gear. It's means that 720k is the AVERAGE dps for your gear, when played optimally. Roughly HALF of the sim results for your gear fall ABOVE 720k. Higher than expected crit rates, nice proc strings, etc all account for more than the "expected" 720k....and guess what? Real life players actually experience those situations in game and their parses go into warcraftlogs. Your parses are measured against those when determining percentile.

To say "boo hoo I don't have BIS legs ergo I'm gimped" is to lean on an imaginary crutch. Good players should be parsing 90-95% even with garbage legendaries.

[Deleted]

You're building the entire basis of your argument on a faulty foundation. Sim dps has exactly ZERO to do with the discussion at hand. Your statement completely disregards the bell curve in SimC's results. That 720k doesn't mean that it's the max possible DPS for your gear. It's means that 720k is the AVERAGE dps for your gear, when played optimally. Roughly HALF of the sim results for your gear fall ABOVE 720k. Higher than expected crit rates, nice proc strings, etc all account for more than the "expected" 720k....and guess what? Real life players actually experience those situations in game and their parses go into warcraftlogs. Your parses are measured against those when determining percentile.

To say "boo hoo I don't have BIS legs ergo I'm gimped" is to lean on an imaginary crutch. Good players should be parsing 90-95% even with garbage legendaries.

Yeah, you did not get it and you have no clue, sorry

1. I compared it to my SimDPS to show that even close to perfect execution meant parsing at the 60th percentile. I have pretty great gear but my DPS is barely average on a simple troughput boss simply because 90% of the players above me have a 2nd good legendary that adds 10s of thousands of DPS.

2. You don't understand the bell curve yourself. 95% of the 100.000 simulated fights fall into a very narrow window (700-740k maybe). Me parsing at the 60th percentile is not due to many players above me just having some great RNG tries. Maybe 20 logs or so.

3. At the very top, where you don't have 10s of thousands of bad players logging heroic kills or easy mythic bosses, you only have good players with mostly great gear. Having good legendaries (plural!) makes or breaks your parses there.

But please, educate me about warcraftlogs, parses and performance some more

Ring got nerfed a long time ago, it's bis but its going to be a small increase over your current legendaries.

Tell me more about my class and the legendary I play with. The ring is still worth ~8-10% DPS which means about 60-70k DPS for my gear level. If I had the belt that would add about 30k over the trinket I use. The shoulders would probably be around a plus of 80k.

[IceMan1763]

First of all, thank you all for your feedback. Let me address some of the points that were made one by one:
Indeed, but I'm not sure if this is currently feasible in a systematic fashion. I'll look into this some more and see what I can find. For now I'm inclined to think that their impact outside of the highest parses should be relatively small compared to the spread induced by player skill. Furthermore, at this point the argument that more people with the best legendaries should be among the highest ranking parses remains an unproven assumtion. There could be just as many bad players with the right legendaries at the bottom half of the distribution.

Personally I'm convinced that people use the argument of not having gotten the right legendaries as a subjective excuse for their own objectively poor performance. I've had plenty of raids where the people who complained the most turned out to rank way below the median when we evaluated the logs. As long as you're not ranking in the lower 90s you should not worry about the impact of legendaries and instead focus on becoming a better player. There are players who do not have the best legendaries among the top parses for every class.

Considering the amount of variance at the top end, and stipulating that we're only interested in players in the top 25th percentile as skilled players are obviously going to parse there even without decent legedaries, I think hypothesizing that the difference between the 75th percentile and 100th percentile is mostly legendary related is reasonable. It's also reasonable to hypothesize that the variance in that top 25th percentile is larger in this expansion than previous ones due to legendaries as well.

[Deleted]

Thank you again for your valuable feedback.

It seems to me that my conclusions could have been misread, so let me clarify a few things.


1) The data shows distributions of DPS, which is not the same as player skill.

These things are of course not the same. The problem is that we can only observe DPS, but not player skill. It's plausible to assume that we'd observe a normal distribution of skill (ability) in large samples of players if we could measure it appropriately, but this remains an assumption. However, the large amount of data that goes into (many, but not all of) these graphs means that there will be many different combinations of luck and skill. Imagine the respective DPS distribution as the combination of a standard normal skill distribution and a spec-specific RNG range. The top parses will always include a combination of flawless execution, optimized gear (including legendaries), and godly RNG. Consequently, the further you look at the right tail the more you will overestimate that player's skill (due to the RNG component) if you interpret the DPS percentile as a skill percentile. Having said that, I've adjusted my statement in the opening post accordingly to

Highly skilled players, under favourable circumstances, can push even the weakest specs into top DPS ranges.


2) Legendary items

On the subject of legendary items, let it be noted that I also think that the system is absolute garbage. It's an additional layer of RNG without the mitigating factor of frequent repetition. It's impactful without being skill-dependent and probably the reason why we're seeing these long right tails. I'm saying probably because the data would indeed have to be normalized for legendary possession to allow us to resolve the argument.

However, we can draw some limited conclusions from the data on mythic difficulty. It's safe to assume that most of these players are talented and very likely to parse high on heroic difficulty. It's also likely that relatively more of them have the best legendaries for each spec because they play more. If you believe both of these assumptions you could go ahead and interpret the differences between the specs in that particular fight. If everybody's good and well-geared, the relative differences between the specs seem to increase in some cases and remain comparable in others. Note, however, that the rankings still change between fights.


3) Balacing a system this complex perfectly is impossible.

One of the best comments that I read demanded that balance discussions should be based on equal skill and gear levels. I'd even add "for the same fight and difficulty" to that and I provided the data on a similar item level because I agree very much. However, it's an illusion to think that this would resolve the debate. To illustrate this, let me use an example where things are a lot closer to this ideal. Can you guess?

It's Arena PvP. People who play frequently are competing with identical gear against people at an identical objective skill level. I don't think people who play PvP under these conditions consider it particularly balanced. What happens is that the discussion of balance just moves on to the next issue in line, which could be team composition or other things. Should you only allow people to compete against similar team compositions? Only with similar specs? Only on one map? The only game that has ever been considered to be nearly perfectly balanced is StarCraft Broodwar, and even there people complained about the impact of the maps.

World of Warcraft is orders of magnitude more complex. Inevitably, you have to draw a line somewhere and defend your choice. It's currently not clear to me where that line is and I certainly do not consider the job of balancing the game done. It can never be. If I did I wouldn't have bothered to provide a detailed breakdown that can be utilized to identify specific interactions of between specs and fights to then investigate them further. However, and I stick to that point, given the complexity of the problem (which they unquestionably increased with the introduction of legendaries) I think they are doing a good job.


4) Meta-interpretation

This is probably a point where I'll remain in disagreement with many people. That's alright. I think at this point everyone can take the data and decide for themselves where they draw their line, but I don't think it makes a lot of sense to discuss arbitrary benchmarks.

What I meant to show instead was that every spec is good at something and all of them, given the right circumstances, can be taken to the top. And none of them is consistently over- or underpowered.

[Raiju]

Tell me more about my class and the legendary I play with. The ring is still worth ~8-10% DPS which means about 60-70k DPS for my gear level. If I had the belt that would add about 30k over the trinket I use. The shoulders would probably be around a plus of 80k.

I play the same class, the ring caps at about 6% now. The shoulders wouldn't even be close to 80k.

[Trend]

Tell me more about my class and the legendary I play with. The ring is still worth ~8-10% DPS which means about 60-70k DPS for my gear level. If I had the belt that would add about 30k over the trinket I use. The shoulders would probably be around a plus of 80k.

There isn't a combined 20% difference (your claimed 60-70k for the ring, 80k for the shoulders) between the legendaries you have and the legendaries you want. I'm sorry, but there just isn't. When you make such absurd claims, it really makes you sound like someone who just can't handle the truth that you're not nearly as good as you want to think you are.

Don't get me wrong, I think the RNG with legendaries is terrible, just don't get stupid with the argument.

[Jimson]

You're building the entire basis of your argument on a faulty foundation. Sim dps has exactly ZERO to do with the discussion at hand. Your statement completely disregards the bell curve in SimC's results. That 720k doesn't mean that it's the max possible DPS for your gear. It's means that 720k is the AVERAGE dps for your gear, when played optimally. Roughly HALF of the sim results for your gear fall ABOVE 720k. Higher than expected crit rates, nice proc strings, etc all account for more than the "expected" 720k....and guess what? Real life players actually experience those situations in game and their parses go into warcraftlogs. Your parses are measured against those when determining percentile.

To say "boo hoo I don't have BIS legs ergo I'm gimped" is to lean on an imaginary crutch. Good players should be parsing 90-95% even with garbage legendaries.

I don't think you understand how Simcraft works. Yes, half the results will fall above 720k... but that's irrelevant. What you SHOULD be asking is "What is the range between the lowest and highest simulated DPS?" I think you'd be surprised at how narrow that range is.

[Zan15]

there are several classes that are clearly outliers on the majority of fights and the fights where they are not its because of padding.

Balance of the dps specs is in a good place, if you are standing still doing nothing but dps.

even then the difference between top and bottom is still too large. 10% on some fights, 5% on most.

[ufta]

I don't think you understand how Simcraft works. Yes, half the results will fall above 720k... but that's irrelevant. What you SHOULD be asking is "What is the range between the lowest and highest simulated DPS?" I think you'd be surprised at how narrow that range is.

I'm well aware of the range, and I don't consider it to be particularly narrow.
For my own DH, the range between the lowest and highest simmed dps goes from 85.3% of the avg to 119% of the avg. That's a 33% variance caused by nothing but RNG. Same player, same variables, same gear. To me, that is not narrow at all.

[satimy]

Class balance now is garbage

[Sarevokcz]

This is a logical fallacy for the high end of performance this expansion. At a certain level of player performance (for the sake of it lets say 95% of the optimum), player skill becomes irrelevant and legendaries become the defining factor. When you progress mythic a a decent level, it doesn't help you a lick that you are able to (going full cliché) outperform a 50-year old housewife even if she has 30 item levels more than you. That is irrelevant for balance discussions which should be based on equal player skill and gear.

Even if I reach my simulated 720k DPS for the Krosus mythic fight (which I did or one of my kills with good RNG a few weeks back) it does not even put me at the 70th percentile. Why? Because I have only one of the 3 best DPS legendaries. You can look up my character, my gear is pretty good. Since I analyze my logs in depth, I can guarantee you that I play my rotation cleaner than almost any other player and as close to "perfect" as you can be. But you can't overcome legendaries making up 10s of thousands of DPS.

It is not a sign of good balance that the absolute best Balance druids with BiS gear and great RNG can beat braindead Demon Hunters or whatever. It just isn't.

btw: saying someone plays their class to 95% (or whatever) of their potential is not equal to them being better than 94% of the other players. Utilization your classes potential is not a linear graph since even bad players do at least 30-40% of the potential DPS. So there are many, many players that are doing pretty good on that front.

first off, krosus is BAD example to base ST damage on, since for some reason, DW classes dont have their usual 20% miss on autoattacks and obviously that means alot more power for specs, that have strong autoattack procs, like demon hunters, its basically 15-20% more fury gen from AA sources, etc. augur is probably more fair

second, if you play your spec very cleanly with 1 good dps legendary, the chances are you are in 70% bracket because
A: you do mechanics to the T or you get targeted by annoying mechanics, which means sometimes sacrificing some dps
B: your raid is changing some tacts, that arent the best for maximizing dps output (saving CDs for specific adds, stopping dps to wait for something etc)
or
C: you realize alot of higher logs already have significant add padding, as in, on krosus they let more adds spawn (preferably in melee) and cleave them through CC for extra bits of deeps

[Deleted]

I play the same class, the ring caps at about 6% now. The shoulders wouldn't even be close to 80k.

PM me your armory link and I will sim your character for you.

I will post sim results for my character without the ring i got and with ring+shoulders. I will take the Star Augur fight /pure ST, ~510 seconds as reference).

This is getting ridiculous.

[Raiju]

PM me your armory link and I will sim your character for you.

I will post sim results for my character without the ring i got and with ring+shoulders. I will take the Star Augur fight /pure ST, ~510 seconds as reference).

This is getting ridiculous.

https://zarania.github.io/zarania/22...gendaries.html

[Deleted]

https://zarania.github.io/zarania/22...gendaries.html

1. Single DPS simulation. People wear 2 of them.
2. Generic gear for all simulations, which ofc hurts/helps certain legendaries more than others (using the shoulders without CoF and 1 Meta-relic is ofc shit)
3. Near BiS gear means more crit/haste than most players will have on live

So pretty much useless.

It is really simple: over a Krosus fight of ~6 minutes the ring generates 800 more fury on average, which translates to about 27 more Chaos Strikes/Anihilations (50% crit chance means ~30 Fury for each one). The ratio of Chaos Strikes to Anihilations is ~3:1 the damage per Execute 750k vs. 1.100k. That equals ~22.700.000 over 360 seconds or 63.000 DPS.

Or you could just look at logs and realize that only ~10% of the top logs are without 2 of the BiS DPS legendaries - the "I Crit like hell during Meta(s)"-outliers.

[Raiju]

1. Wearing two does not drastically impact their increase, other than for a special case with the shoulders (and even then)
2. We both know the ring is more important anyway, and it isn't gear dependent.
3. Most people are over 900 at this point on their active main, many above 905/910. Also with more gear you gain MORE dps from legendary effects, not less.

I never even implied these aren't the top legendaries (well, shoulders isn't) - just that you are DRASTICALLY overstating the benefit they'd give you.

[Deleted]

It's not "just as bogus".

We have no good way of distinguishing high parses that happened due to players being skilled and high parses that happened due to random procs / artificially created special circumstances. If we had a good way to do that, we would have filtered out all of the latter and then we could proceed to analyzing what's left. It would still make sense to analyze the middle and the ends (both high and low) separately, but at least the ends would be analyzed.

But since we have no good way to filter out all of the fluff and the history shows that quite a lot of outliers indeed happen because of random procs and things like that - take 2-3 top graphs for any boss and I bet at least one of them - quite possibly all of them - will contain things like that, we can either filter heavily and try to analyze the averages, speaking about the results as those for averages, or we have to forfeit the analysis completely because we already know the outliers are skewing it heavily and tons of them are artificial.

It is just as bogus. The methodology is at least. To disregard all extreme values, outliers and valid alike, reduces the variance beyond what we know is reasonable. Your conclusion might be closer to the true situation of spec balance, but it would be purely coincidental if that was the case since we don't know if the problems generated by outliers outweigh the problems of missing (removed) data.

We could 'just' make a set of conditions an observation should not violate to be included in the data set. It isn't rocket science, but it is tedious. It could be something like a condition on uptime on buffs/procs pr. minute to account for an internal spec situation where it is unreasonably high. You could also put up restrictions on "damage taken by " or even "damage dealt to " to filter for tactics skewing dps, like on Skorpyron or Tich, if that was wanted to test the hypothesis. It is definitely not impossible to make a reasonable filtering compared to the arbitrary one to cut-off at x and 1-x percentile. It would require a lot more work, which I assume was your point with "we have no good way", as it might not be feasible unless you have a lot of spare time on your hand to spend on a hobby.
But just because it might not be feasible doesn't make an analysis disregarding the tails better. It is still not good since the number of explanations for differences in dps-output between specs for the middle of pack of players could be endless and not just the spec; gear optimization, skill, legendaries, tactics, unrealistic circumstances (should also exist for worse players that don't make it over the 75-percentile), and probably even more, I haven't considered, could also be (part of) the explanation.

[Grmmppff]

btw a feral(if you still manage to find one anywhere), and I guess a few other specs too, would disagree with the all specs are viable nonsense. All CLASSES do in my opinion have a more or less viable spec that can perform near the top, and this would be great, if a) there was no such thing as traits(or if you had 54 traits in all your specs) and b) legendaries improved a class and not a spec. Fire mages now feeling forced to go frost, druids feeling they have to be balance(or resto/guardian,which has nothing to do with dps, just to show druids(THE CLASS) are in a very good spot atm), dks feeling they need to be frost and breath-specced(and I'm sure there's other classes,those are the ones I play), makes for other specs(and their legendaries) feeling like a huge waste of time

[furiousferret]

Anyone that's complaining about class balance probably doesn't remember the times where certain classes were literally useless in certain patches, and it took months to fix it. I remember when enhance was one of the worst specs (end of bc or wrath?) and they still nerfed our damage, claiming it was needed to be done to fix it. That fix didn't come for a long time.

There are weaker classes now, but the gap is really narrow comparatively.

[Riistov]

It would be interesting to see how much the proper legendaries affect dps ranges.

Agreed! They seem to create a huge variance in the case of many specs.

[Infestor]

please don't nerf warlock.