Guild Wars 2

I would like to get into a scientific analysis of our skills compared to that of other classes. For this evaluation to be meaningful we would have to include all factors affecting a skill.

Let’s take an example:
Elem’s lightning whip versus warrior’s axe chop chain.

• coefficient per second: 1.22 LW vs 1.36 Chop
• range: 300 LW vs 170 Chop
• commitment: must not interrupt chop’s chain to deal its damage versus uncommitted LW

So the question for these skills is as follow: are the extended range and commitment freedom enough to justify the loss in damage?

It is almost impossible to answer this question without further data. This is why I am asking for your help: I would like the community to do comparisons similar to what I have just showed you in order to establish the “price” of range in terms of damage, for example.

I am actually planning to use regression analysis in order to accurately determine the “price” of all factors in terms of damage

Please help me settle this debate scientifically!

All I know is that LW is much more beautiful. It’s a classy, flashy, and sadistically exquisite way to inflict pain.

Zelyhn:

Its a nice initiative, and appreciate you doing this so we can have a more constructive discussion.

However:

1. We should wait until the 15th balance patch until we know if and what numbers are changed.
2. When taking into account damage and range, you should also take into account armor which you made a topic about: https://forum-en.gw2archive.eu/forum/professions/elementalist/Numbers-are-fun Because armor determines how long you can do sustained close range dps. One ele does not simply infinite dodge everything.

I hope this thread is getting the comments you asked for ^^

Coefficients of autoattacks don’t show whole picture.

Thanks Wiz

Yes I should wait until the 15th but I am in a 5 hour train and I am awfuly bored ^^

As for amor, I think it’s the point of this discussion to not include it, along with traits and class mechanics, in order to evalutate the skills in themselves not the class as a whole. If I was to include armor in my evaluation I would also have to include many many more factors specific to the classes which are hard to measure.

Carrying on: arc lightning (scepter AA) versus spatial surge (mesmer GS auto)

• coefficient per second: 0.47 for AL versus 0.38 for SS (on average over full range), 24% difference
• max range: 900 AL versus 1200 SS.
• comittment: high on AL, low on SS

You cant just compare 2 skills, let alone just 2 aspects of it.

A warrior has infinitely better ways to sustain itself and survive in melee than an ele could ever hope too. They have higher hp and higher armor bases, healing signet which is by far the best mindless sustain in the game, hands down. A warrior will be able to stick to you like fire in a dry forest, whereas the ele is being constantly being poured water and cut possible paths.

D/D should be the stronger melee weapon given how little survival we have at melee range. But this isnt D/D problem’s only. We are probably the most unsafe profession out there, even at max range with staff. No disengages, no damage, no healing, long CD on the only respectable skills, whereas the equivalent in some other classes have 10-20 secs CD at max, we have 30-40, on top of 12-15 secs attunements.

I dont want to be gloomy guy but for every single thing we can do better than any profession, that same profession can do 2-3 things better than us.

haviz.1340:

Coefficients of autoattacks don’t show whole picture.

I agree but “showing the whole picture” is a task that requires many preliminary steps in order to be accurate. I am just making one of these steps here.

Good luck! I don’t even know the exact hit areas of all moves so i’d be hopeless at that. I do know that both LW and Axe chop are a class of moves called ‘Cleaves’ by anet. I assume every move falls into a category like that. It could be that every Cleave has the same hitbox (i assume a cone in front of you). I’d check if that’s the case before you waste your time comparing apples and pears.

Zelyhn.8069:

I am actually planning to use regression analysis in order to accurately determine the “price” of all factors in terms of damage

Please help me settle this debate scientifically!

Just curious, what is the dependent variable in your regression? Sustained damage in a real fighting scenario? If that is what you want then you will need to model many factors (like mentioned before, toughness, survivability should also be considered as factors, as it does not matter how hard you can hit, if you cannot survive, your sustained damage will still be low).

A simpler way to model this is to evaluate a build’s sustained damage under certain situation instead of modeling specific factors such as damage coefficient of a skill and toughness. For example, you can compare S/D’s sustained damage in a boss fight, compared to a glass cannon warrior build. Of course group composition will also influence the sustained damage. You may need to take lots of samples and then do stats analysis.

If you want to do this using science, you have to consider the trade off between internal and external validity. You can design a very controlled study, but due to its highly controlled nature, it has little generalizability to real game play. Or you can design a more realistic study, but your ability to do causal inference may suffer.

In my own experience, eles can do great damage, but these builds are usually not very viable in the current pvp meta. As for pve, since it is not competitive, it really does not matter as much.

Guys please try to understand that I am not trying to solve the debate of which class is better. So no, I won’t do anything “realistic” because way too many factors would have to be included. I want to stick to comparing skills. From this comparison we may conclude something on class balance, but this is a later step.

@Illusion
Yes the dependent would be damage, expressed as a coefficient per second.

While not damage so maybe a little off-topic, this is still my favorite example of skill comparisons: Windborne Speed versus Charge

Here is some old data that might be helpful:

http://www.guildwars2guru.com/topic/49249-weapon-skill-coefficients/

I would say, try to compare a full weapon to one another (for instance air attunement vs. axe) as some weapons have better autoattacks (ranger SB) and 2-5 abilities that are more or less effective. For our own air, that autoattack is good, but the rest of abilities deal almost no damage (with the exception of RtL, which is only useful for slightly more burst but best saved to get out).

dukevonart.2691:

While not damage so maybe a little off-topic, this is still my favorite example of skill comparisons: Windborne Speed versus Charge

Five times less range and twice longer cooldown

Fortus.6175:

You cant just compare 2 skills, let alone just 2 aspects of it.

A warrior has infinitely better ways to sustain itself and survive in melee than an ele could ever hope too. They have higher hp and higher armor bases, healing signet which is by far the best mindless sustain in the game, hands down. A warrior will be able to stick to you like fire in a dry forest, whereas the ele is being constantly being poured water and cut possible paths.

So true. Which is why I don’t really understand your next comment:

D/D should be the stronger melee weapon given how little survival we have at melee range.

We have quite a few tools for this. We can get perma protection. We have heals in water. We have Shocking Aura. We can step back a bit from the bosses and not get hit so much while still hitting them. We too can use a on-hit signet.

The weapons aren’t easily compared. In fact, given that Lightning Whip hits twice, are you absolutely sure it deals less damage?

Two skills can be compared, I don’t see why not.

Yes I take the two hits of LW into account for the 1.22 c/s figure.

Let’s try not to get into the class VS class debate but rather in skill VS skill

Before you continue, I want to ask you: In all seriousness, what is the point in digging into the damage difference between skills?

What is the aim, the overall goal, of this post?

Zelyhn.8069:

Two skills can be compared, I don’t see why not.

Because the result is meaningless when you consider the fact that D/D eles only use their auto attacks when everything else is on cooldown (about 10% of the time) and warriors use them a lot more often (at least 30%).

Comparing single attacks for dps is meaningless because when considering the whole build, the ele’s dps becomes a LOT higher.

Yes I take the two hits of LW into account for the 1.22 c/s figure.

Did you also consider their actual casting times + skill delays?

The title of this thread is “are our skills really weak?” not “is our dps that low?”.
It’s a specific question. When I want to work on dps, like I have done in the past, I make it clear. Here I am comparing skills of other classes to ours, in order to determine if our skills are weaker. Why would I include anything else but the skills in the comparison ??

I am using data for cast times + animations to the best of my knowledge. There are actually quite a number of people data-mining for the ele at the moment, if I did a mistake in one of my numerous posts it will be noticed.

Zelyhn.8069:

The title of this thread is “are our skills really weak?” not “is our dps that low?”.
It’s a specific question. When I want to work on dps, like I have done in the past, I make it clear. Here I am comparing skills of other classes to ours, in order to determine if our skills are weaker. Why would I include anything else but the skills in the comparison ??

That’s all fine and dandy, but you’re going at it the wrong way. Let me give you an example:

If I have 2 builds of 2 skills each, that each have the same attack speed on their #1 and #2 skill respectively:

Build 1’s auto attack deals 10 damage and his second skill deals 100 damage every 5 seconds.

Build 2’s auto attack deals 9 damage and his second skill deals 500 damage every 5 seconds.

If you’d only compare the first skill, you’ll see build 2 has a weaker auto attack. But if you compare the whole build, you’ll suddenly see that build 2 deals a LOT more damage than build 1.

Yep but that comes at a later stage

No harm in comparing skills slot by slot for now, no serious conclusions have to be drawn from that, but it is interesting as a start

Zelyhn.8069:

Yep but that comes at a later stage

No harm in comparing skills slot by slot for now, no serious conclusions have to be drawn from that, but it is interesting as a start

You’ll not draw serious conclusions from it, but some people will. We’ve seen what a warrior 100b movie on some orr npcs can do for the community’s view on dps.

Ok let it be said then: I am the only one abilited to draw conclusion here

I can live with that restriction. :-)

hes talking about PVE DPS he doesnt care about anything BUT dps… if im wrong then yea no point in comparing this (BUT IM NOT)

Let’s say the next patch adds a playable turtle class that has one skill. A cannon that does 50k damage is mounted on its back. The only downside is that every other class can one-shot the turtle, and the turtle moves at turtle speed. Now, if we do a dps comparison, then are we having the right conversation?

tldr – Look! I can make hand puppets!

You just picked the best skill in D/D Eles(LightningWhip) Among the other 19 skills.
Try yo compare VaporBlade, Impale

Also, since you are comparing to Warrior,
did you ever,ever consider about the 8000hp and 2000armor differences???

ImProVocateur.5189:

Let’s say the next patch adds a playable turtle class that has one skill. A cannon that does 50k damage is mounted on its back. The only downside is that every other class can one-shot the turtle, and the turtle moves at turtle speed. Now, if we do a dps comparison, then are we having the right conversation?

tldr – Look! I can make hand puppets!

Yeah we are because the only place where dps is worth calculating is PvE and in such mode no player will attack the turtle.

Oh but wait i’m not even doing a dps comparison … !

Now can we get to comparing skills or people want to keep asking whether I take class hp, amor dyes and hair style into account ?

Zelyhn.8069:

ImProVocateur.5189:

Let’s say the next patch adds a playable turtle class that has one skill. A cannon that does 50k damage is mounted on its back. The only downside is that every other class can one-shot the turtle, and the turtle moves at turtle speed. Now, if we do a dps comparison, then are we having the right conversation?

tldr – Look! I can make hand puppets!

Yeah we are because the only place where dps is worth calculating is PvE and in such mode no player will attack the turtle.

Oh but wait i’m not even doing a dps comparison … !

Now can we get to comparing skills or people want to keep asking whether I take class hp, amor dyes and hair style into account ?

hp, armor, DYES and HAIR STYLE? Seriously, man?
If you really want to compare skills ONLY, then you are not discussing balancing, you are just talking about SKILLS

Let me give a picture to you, when you is a SWAT, would you go in to action with:
1- a MP5 subMG or
2- a MP5 subMG + a kevlar vest ?

Well I am just talking about skills
I will discuss balancing later, when I have more information about the relative strength of our skills.

Zelyhn.8069:

Well I am just talking about skills
I will discuss balancing later, when I have more information about the relative strength of our skills.

skill wise we’re pretty good. We got several good hard hitting attacks (provided the target stands still for long periods) we got a good number of condition types. We got a number of CC skills. we also have 3 attack range styles Dagger= melee, scepter= mid, staff=long.

There has been a recent change on the wiki page for the warrior’s axe chain. The coefficient per second of that chain seems to be 1.46 as opposed to the 1.36 previously used.

However, the guardian’s sword auto chain has a coefficient per second of 1.24 and a range varying between 150 and 300 for a commitment time of 2.5s. It seems to me this is slightly worse than LW.

Mesmer’s sword auto chain has a coefficient per second of 0.96, applies two stacks of 5s vulnerability and rips a boon for a commitment of 2.5s and range is 130.

To be honest what I expect to find from this study is that your first 2 or 3 skills in each attunement are equivalent to that of other classes, and that skills 4 and 5 are balanced with longer cooldowns. This would make sense to me: since we have 4 attunements we can have 4 sets of long cooldowns running at the same time versus two for other classes so our skill’s long cooldowns should be about twice as long as that of other classes (actually a bit less to compensate for the constraint of attunement cooldowns).

We have weapon swap on 16s cooldown and no other profession mechanic like adrenaline, virtues, toolbelt, shatters, deathshroud, pet or even stealing. Our autoattacks except lightning whip and fireball should not be compared to other classes because they have other tools which ele lacks or is worse at.

Mesmer dps from weapons is probably the worst in the game, closely followed by ranger. Guess why?

I just noticed that attunement swap cooldown is 2s, not 1 like I thought, and can be reduced by points in arcana … woaw … I have to rework all my rotation TC now.
I had a feeling that attunement swap cd was too high, especially when running D/F persisting flames + FA, because it requires fast swapping … obviously the game does not allow that. My disappointment is high as you can guess by the numbers of “…” I am using …

To reply to you, I agree, except that engineer’s toolbelt is actually its utility bar and kits are their attunements. But that’s for engis; they have other issues.

haviz.1340:

Mesmer dps from weapons is probably the worst in the game….

I don’t think this is true, mesmer just doesn’t rely on his autoattack to do his dmg, that’s all. Well played mesmer can dish out very respecable amount of damage, trust me

STRanger.5120:

I don’t think this is true, mesmer just doesn’t rely on his autoattack to do his dmg, that’s all. Well played mesmer can dish out very respecable amount of damage, trust me

They have less damage in their weapons because good chunk of their dps includes phantasms.

STRanger.5120:

haviz.1340:

Mesmer dps from weapons is probably the worst in the game….

I don’t think this is true, mesmer just doesn’t rely on his autoattack to do his dmg, that’s all. Well played mesmer can dish out very respecable amount of damage, trust me

Bloody hell learn how to read!

No, of course not ! Why would you think that way?
Eles have the most op skills in the game ,such as fiery greatsword, ice bow, and crazy healing ability. And these skills benefit EVERYONE in your party.t

haviz.1340:

STRanger.5120:

I don’t think this is true, mesmer just doesn’t rely on his autoattack to do his dmg, that’s all. Well played mesmer can dish out very respecable amount of damage, trust me

They have less damage in their weapons because good chunk of their dps includes phantasms.

Exactly, so any comparison of skills itself is really pointless, because of class mechanics.
Only skills that can be compared to each other somehow are utilities (blink vs Lightning flash f.e.) And even then, the other options of the certain class should be taken in account…
@OP: Sorry, but this discussion is hardly leading anywhere, the game mechanics are too complex for that.

The fact that some classes’ mechanics are superior to other does not matter … I don’t know why it is so hard to understand that comparing skills is relevant in the sense that it is a first step to evaluate class balance.
I’m loosing my nerves a bit today after learning about attunement swap cd being at 2s.

STRanger.5120:

Exactly, so any comparison of skills itself is really pointless, because of class mechanics.
Only skills that can be compared to each other somehow are utilities (blink vs Lightning flash f.e.) And even then, the other options of the certain class should be taken in account…
@OP: Sorry, but this discussion is hardly leading anywhere, the game mechanics are too complex for that.

It’s not pointless because ele has no mechanics different from warrior.

EDEN.2061:

No, of course not ! Why would you think that way?
Eles have the most op skills in the game ,such as fiery greatsword, ice bow, and crazy healing ability. And these skills benefit EVERYONE in your party.t

Fgs is going to get fixed and ‘crazy’ healing ability is not crazy and is completely useless in pve.

haviz.1340:

STRanger.5120:

Exactly, so any comparison of skills itself is really pointless, because of class mechanics.
Only skills that can be compared to each other somehow are utilities (blink vs Lightning flash f.e.) And even then, the other options of the certain class should be taken in account…
@OP: Sorry, but this discussion is hardly leading anywhere, the game mechanics are too complex for that.

It’s not pointless because ele has no mechanics different from warrior.

I have a bad news for you, if you really believe that, you’re playing your Elementalist wrong. Sorry.

Zelyhn.8069:

The fact that some classes’ mechanics are superior to other does not matter … I don’t know why it is so hard to understand that comparing skills is relevant in the sense that it is a first step to evaluate class balance.
I’m loosing my nerves a bit today after learning about attunement swap cd being at 2s.

It’s not hard to understand, but it’s not a step to evaluate class balance, because the output of this analysis, the result will be total non-sense. You can’t just take one aspect of the class, analyze it separately and hope to get a valid result.

STRanger.5120:

Zelyhn.8069:

The fact that some classes’ mechanics are superior to other does not matter … I don’t know why it is so hard to understand that comparing skills is relevant in the sense that it is a first step to evaluate class balance.
I’m loosing my nerves a bit today after learning about attunement swap cd being at 2s.

It’s not hard to understand, but it’s not a step to evaluate class balance, because the output of this analysis, the result will be total non-sense. You can’t just take one aspect of the class, analyze it separately and hope to get a valid result.

Yes I can.

STRanger.5120:

haviz.1340:

STRanger.5120:

Exactly, so any comparison of skills itself is really pointless, because of class mechanics.
Only skills that can be compared to each other somehow are utilities (blink vs Lightning flash f.e.) And even then, the other options of the certain class should be taken in account…
@OP: Sorry, but this discussion is hardly leading anywhere, the game mechanics are too complex for that.

It’s not pointless because ele has no mechanics different from warrior.

I have a bad news for you, if you really believe that, you’re playing your Elementalist wrong. Sorry.

Yeah Haviz is widely known for being bad at playing ele.

I’m sooo fruuustrated today!!

Edit: seriously Haviz, really, l2p!

STRanger.5120:

I have a bad news for you, if you really believe that, you’re playing your Elementalist wrong. Sorry.

Please explain then, what unique mechanic has elementalist except 2 more weapon swamps on longer cooldowns that warrior does not have?

There are no pets, clones, phantams, minions, death shroud, aegis, adrenaline and reflections are bad at best. Only reason why ele sucks is because it’s supposed to be jack-of-all-trades, problem is, hybrids never work. They are either too strong or too weak.

haviz.1340:

Please explain then, what unique mechanic has elementalist except 2 more weapon swamps on longer cooldowns that warrior does not have?

There are no pets, clones, phantams, minions, death shroud, aegis, adrenaline and reflections are bad at best. Only reason why ele sucks is because it’s supposed to be jack-of-all-trades, problem is, hybrids never work. They are either too strong or too weak.

Hmm, you really don’t see a difference between a simple weapon swap and attunement switching?
(your clue is written in bold FYI)

Zelyhn.8069:

STRanger.5120:

Zelyhn.8069:

The fact that some classes’ mechanics are superior to other does not matter … I don’t know why it is so hard to understand that comparing skills is relevant in the sense that it is a first step to evaluate class balance.
I’m loosing my nerves a bit today after learning about attunement swap cd being at 2s.

It’s not hard to understand, but it’s not a step to evaluate class balance, because the output of this analysis, the result will be total non-sense. You can’t just take one aspect of the class, analyze it separately and hope to get a valid result.

Yes I can.

Ok, I’ll wait for the results then