Showing Posts For Azzumy.7685:
Outgoing healing power is multiplicative.
I tried healing with Guardian, Elementalist, Tempest, Warrior and Revenant. Never tried Necromancer, because with the idea that Necromancer would fail at it I never looked into it.
On most classes (except warrior) I achieved at max 2k heal per second, so the healing you calculated here would be enough as main healer.
The main thing I noticed is that heal per second is not saying how good it is at healing. The chance to hit (or radius) is a lot more important. For example the worst healer I have is the elementalist, because water staff auto attack heals a lot, but the chance is very small you will hit a party member with it. The best Healer is Tempest because the shouts have 600 radius, but the heal per second is not optimal. My main healer before HoT was guardian, it has around 300 radius heals which is really the minimum size.
Sigil of renewal,360 radius, 183 h/s
Life from death 360 radius, 715 h/s
Transfusion, 600 radius, 162 h/s (assuming per pulse)
Well of blood, 240 radius, 232 h/s (Annoying light field)
Mark of blood, 240 radius, 547 h/s (permanent regen)
When you hit everything it heals at max 1709 h/s.
Skipping all radius 240 skills then you are at 1060 h/s.
For PvE you have to remove runes of Monk and rice ball to do condition damage, this will reduce the total healing to around 800 h/s. However the burst heal is good, but you are better off using Ritual of Life (20% revive speed) and Runes of Mecry (revive at 80% hp and revive speed).
If you want to support it is better to use tempest, this can give protection and frost aura, which is reducing their damage taken by a factor of 2. With the healing build you would increase their healing with a factor of 2. In dungeons the only thing missing is either stability or CC. For raids, during the Vale guardian you can prevent most damage by knocking the seekers backs.
If you want to support, then boons is the way to go. With this build you double their healing, with a tempest build you can reduce their damage taken by a factor of 2 (Protection+frost aura). Or in most dungeons the only thing missing is stability or CC. In the raid at Vale Guardian you can prevent most of the damage by simply using knockbacks.
Should Anet be looking at this?
In general full healing power (1800) should increase healing by ~90%, to be equal to toughness.
(edited by Azzumy.7685)
- If you are playing a condition build you will have to pick Scepter as your main weapon
- If you want Dhuumfire to do anything you will have to sit in Shroud auto-attacking for a significant amount of time
- If you sit in Shroud auto-attacking you lose out on condition application from your Scepter, off-hand and Utility skills
I have a conditon build with dagger/focus + greatsword. Outside of Reaper shroud all I do is recharging life force (on this way I have a Reaper uptime of around 80%+). So your first point is not valid.
Dhuumfire is the main trait that gives diversity to condition builds. This means that you either use Dhuumfire and sit in Shroud auto-attacking for a significant amount of time, or you use scepter. I know one hybrid type of build that can use both at ~90% efficiency.
The goal with Shroud and Dhuumfire is that you make sure all your utility/weapon skills are on cooldown before entering Shroud, on this way your third point is also not really valid.
On top of that, Dhuumfire is pretty much the only reason you might want to take Soul Reaping as a condition Reaper, survivability concerns aside maybe. On the contrary, it makes you lose out on the support and utility from Blood Magic, or (if you are not playing in optimized groups e.g.) the self-might and vulnerability stacking from Spite.
Blood magic has some really nice utility. Having problems deciding between nice traits from Soul Reaping and traits from Blood Magic is what gives Reaper build diversity. Just as that Spite is really nice when soloing, but completely useless when you as condi party up with a Phalax warrior.
Since I play Reaper I am swapping trait lines a lot more based on the situation. I feel like 2 trait lines are required for my play style and the last trait line I pick based on the situation. Currently this works because Soul Reaper has everything Shroud requires, Curse has everything scepter requires and Spite has everything power requires. The last 2 trait lines have nothing that is “forcing” you to pick it, but they do have some really nice traits.
LB is a great weapon to rotate through with quickdraw, but isn’t a full replacement imho due to requiring distance to make the most of it and missing out on party stuff like boons, buffs and condi clears.
The damage from long bow is only slightly better at range compared to melee range, because most of its damage comes from Rapid fire and Barrage. This means missing some auto attacks due to dodging is a neglectable damage loss. However having to cancel Rapid fire or Barrage in order to dodge will have a massive effect on the damage.
The calculations I did in my first post were done for melee range and in this case bow was as strong as sword. So even if auto attack has more damage at longer range, it does not mean longer range is required.
This morning I started to play on my ranger again and I am wondering why the most common rotations I find are using sword and axe.
Barrage -> Rapid fire -> Auto attack sword +Axe when off cooldown till enemy is dead.
For a situation that requires burst (<30 seconds) I understand this, but in long fights (Raids) it seems that longbow is nearly as good as sword.
Sword and Axe
Sword has a 1.9 power coefficient over 1.8 seconds
Path of Scars is 2.4 power coefficient over 1 second and 15 seconds cooldown (Not sure about time).
So this means that sword and axe has 1.145 power coefficient per second average.
(1.9/1.8 * 14/15 + 2.4 * 1/15)
Predators onslaught makes this 1.26 power coefficient per second average.
Longbow
Longbow with “Lead the Wind” and “Quick draw” does:
0.7 power coefficient in 0.9 seconds auto attacking.
4.8 power coefficient over 2.5 seconds with barrage on 9 sec cooldown.
3.75 power coefficient over 3.3 seconds with rapid fire on 9 sec cooldown.
So this means that Longbow has 1.23 power coefficient per second average.
(4.8 /9 + 3.75 /9 + 0.7/0.9 * (9 – 2.5 – 3.3)/9)
Quickness
Because bow damage is limit by cooldowns there is a larger difference caused by quickness.
Sword + Axe =1.26*1.5 = 1.88 power coefficient
Longbow = (4.8 /9 + 3.75 /9 + 0.7 * (9 – 2.5/1.5 – 3.3/1.5)/9*1.5) = 1.55 power coefficient
Point of discussion
From these calculations, quickness would be the only reason to pick sword over longbow. Assuming my calculations are correct. When ignoring quickness I see little reason to prefer one weapon over the other.
So my question is: Is there something I missed in the calculation or in my search for Ranger builds? Or is there another reason why people prefer one weapon over the other?
Sources
coefficients and sword chain time:
https://wiki.guildwars2.com/wiki/List_of_ranger_skills
Barrage and Rapid fire channel times were determined by:
10 auto attacks in 9 seconds
5 auto attacks + barrage + 5 auto attacks in 11.5 seconds
5 auto attacks + Rapid Fire+ 5 auto attacks in 12.3 seconds
(edited by Azzumy.7685)
..the differences in DPS of the same build, played by the same person… with the same team comp… based on encounter, are so great… that it literally renders all math spreadsheets meaningless even for comparison purposes. …
Math itself is exact and cannot be wrong, however you have to neglect certain aspects and make assumptions in order to simply the math. The conclusion people draw based on this math can be wrong, because these people are not within the boundaries of the assumptions (Human error) .
I assume you would agree that most of the current spreadsheets are complete enough to give the exact damage per second of direct damage dealt by auto attacking Target Golem (Indestructible). But next you say that when everything remains the same except the encounter, that this math is wrong. However if you insert the new armor value and assume no downtime due to stuns or blocks, then this spreadsheet should be exact for direct damage until the time you die. Ofcourse the spreadsheet now only works for the case where you do not avoid damage, but that does not make the spreadsheet wrong (except for scholar runes and some traits).
If you want to start a discussion based on spreadsheet damage calculations, then you should give a very specific spreadsheet and a very specific goal you try to achieve with that spreadsheet. Based on that it should be possible to expand the spreadsheet to make it complete.
An example goal could consist of:
Input:
Character stats (may be variable)
Target health assumed infinite for direct damage (power damage)
Target armor
Necromancer Dhuumfire build (Easy to implement rotation)
Dagger as weapon.
Rotation: Use skill 1 AND go in reaper shroud when life force is full.
Damage uptime (variable) (simplification for dodging, healing, blocks, CC)
The damage the player recieves (as function of time and for given amount of armor)
The healing the player recieves from outside.
Buffs and debuffs provided by party (if any).
Output:
Damage per second averaged
Ramp up time for conditions averaged
Optional:
Scholar uptime estimation
When using farming as argument, warrior has acces to 2 leaps (sword 1 and bull rush) and 2 charges (greatsword). PvE wise both are in a very good spot and have similar damage. Guardian can support with stability and healing breeze, Warrior can support with might and banners.
Depends on what you want. If you want to be most optimal, just spam dagger 1 with sigil of bloodlust and runes that give power.
I would suggest that you pick stuff that is not optimal, because at level 80 the chance is high that you will run a more optimal build and even forget about all the stuff the class has. A not optimal bould would be using axe mainhand with dagger offhand. Or using staff as main weapon. For trait lines, use that what you unlocked last.
Healing is extremely useless in this game, because reviving is a lot more effective and does not give any dps loss. In general people will dislike you when you use healing gear. On the other hand, people will thank you when you revive them. This can be greatly improved with the runes of mercy and the revive speed trait.
You are engineer? The elixer R is a great skill to revive other people, our yourself with a bit of practice. The stealth and reflect from elixer S can also be useful.
Gearwise you could go for boon duration (concentration) (if it is not bugged anymore) and give people around you might, by blasting in fire fields.
Conditions ramp up linear. This means that the ramp up damage loss is half the ramp up time multiplied by the DPS.
9k dps and 6 seconds Dhuumfire duration results in 0.5 * 9k * 6 = 27k damage loss per start or invulnerable phase. Or easier said, the fight would effectively take 3 seconds longer.
This is assuming that all buffs do not have a ramp up time, which is a wrong assumption. So in practice the fight would take around 2 seconds longer due to ramp up time. This is increased by 2 to 3 seconds per invulnerable phase.
Also keep in mind that gravedigger has a 7 seconds cooldown if you hit a boss that just went invulnerable, even if it is just a single block or evade.
… I really liked how you specifically mentioned all of your assumptions, but you also have to realize they are dangerous assumptions to make. …
Almost all these assumptions have effect on the power builds. The raid (as of beta) was during day time. Fury has more effect on power builds. Scholar runes (GS build) requires 90% hp, while the other builds only require a burning target.
The biggest assumption made in the spreadsheet is that all builds have to same uptime. Which depends a lot on the encounter and player skill.
The problem is see is that Reaper’s might has as condition that you have to be in shroud. In order to make shroud a viable option you need the Reaper trait line and the Soul Reaping trait line.
Should there be 3 trait lines that affect the Shroud so much?
Currently with completely skipping spite we are selfish and not even self sustaining anymore.
Hmmm… If the duration of the might from chilling victory would become 15 seconds, then there is also finally a difficult choice between this and decimated defenses. And that would lower the need for Spite for shroud builds.
Ghost pepper popper seems like a bad idea, because against a single target is only gives around 4 stacks on might. The best power food gives 100 power and 70 ferocity, which is nearly 175 stats or equal to 7 stacks of might.
Reaper’s might (Spite trait): Shroud skill 1 grants might.
In my opinion this is one of the main aspects that make Necromancers selfish and is probably one of the most important traits this class has for shroud builds. But this trait is located in the Spite tree, which I think is removing any build diversity.
With the release of Reaper shroud, the Necromancer’s shroud finally started to deal some damage. However this requires the Reaper trait line and the Soul Reaping trait line (Dhuumfire and Vital Persistence). Because of this only one choice of trait line remains. However the might from the spite trait line alone already gives around 45% damage increase at level 80, making it a very bad choice to pick any other trait line.
Would it be better if Reaper’s might would be placed somewhere else? My idea would be to make it base line. Either having reduces base duration and that the spite line increases might duration. Or not changing its duration, but applying self vulnerability. The spire trait line can then have a stacking buff based on the number of vulnerability stacks on yourself. Or a buff based on the number of vulnerability stacks that have been cleansed.
In here is a spreadsheet for DPS calculations.
https://forum-en.gw2archive.eu/forum/professions/necromancer/Reaper-dps-builds
Is there any reason why you use Shroud skill 4 over skill 1? Instead of skill 4 you can do 4.5 times skill 1. This gives more or similar damage from DoT, more direct damage, 4.5 might stacks, 9 stacks of vulnerability and it also generates 2% to 6% life force depending on the number of targets.
How important is the damage outside of Reaper shroud? The time outside of shroud I normally spend on obtaining life force from all non-auto attack skills, healing my health, boon stripping, reviving allies and resummoning minions. This normally already takes up more then the 10 seconds cooldown, not leaving any time to deal non-shroud damage. The only time when I am outside of death shroud is when I am too lazy to keep track of my life force, which I am for most of my time in the open world.
Also keep in mind that Soldier and Dire more than doubles your shroud uptime compared to sinister and berserker. Which might be a downside if you have to boon strip for your party, as I noticed in the fractals 40 to 50.
Before switching to condition, first think about the armor of the enemies you are going to fight. During the cliffside fractal I notice that enemies have very high armor, making conditions ideal. But during the beta the guardian in the raid has very low armor. There are some objects that can that critical damage, but do not take condition damage.
There is video evidence of gravediggers going up to 33.5k against Vale Guardian. It’s in LOD’s raid footage. Lots of hits for 28-29k too.
Which, given Gravedigger’s total cast time of 1.7s, pushes our DPS quite a bit when the enemy is below 50%. Likely not enough to save the class for PvE but it’s something.EDIT: The hits against the legendary wyvern shouldn’t count for much. It takes massive damage when the breakbar is broken.
33.5k damage per hit would mean that the Vale guardian has low armor. Because as far as I know I have taken into account every possible damage booster. And on top of that, sigil of night and dungeon consumable do not even work in the raid. So a raid boss with low armor (probably ~2200) is good news for power builds.
I updated the GS build for correct food and death magic power bonus instead of soul reaping line. I checked Gravediggers total cast time + after cast and it still is 2 seconds.
I could go and defend my other builds, but power builds win against low armored targets.
Edit: Current GS build:
http://gw2skills.net/editor/?vRAQNAodRn0ICV2gt0AO2Ac4gFcBTKH01CbxXxvYZEaAEAaBA-TBSBAB6pJwS1fua/BsU+toSQ80DoxpAAgnAApAiYrF-e
The spreadsheet shows the damage without taking into account any toughness party buffs (banner or guardian aura), which would add ~2% crit damage and 24 power. But it does take into account the buff from corrupter’s Fervor and Flesh of the Master (4 minions).
(edited by Azzumy.7685)
Greatsword dps already went to 12.3k DPS (2600 armor) after some tweaks. Keep in mind that this is with 25 might and 25 vul. Which is impossible to stack by yourself without Reaper shroud. For most bosses I also notice that 2600 armor is too low.
On top of that is a greatsword build squishy. It does not build any life force and with the necromancer healing you also won’t survive or you lose DPS due to dodging. (Might compensate for limited death shroud uptime.)
If you don’t trust the calculation, try setting the stats and damage bonuses for heart of the mist settings and then compare the equal damage to the calculation.
Did some testing of enemy armor in Verdant Brink and Auric Basin.
Raptors, mushrooms, jungle grub have around 2600 armor.
mordrem grant, Stoneheads and boars have 2900 armor
beetles (from front) 4200 armor
Beetles (back) and arrowheads have 6200 armor.
This means that power builds might also want to increase the enemy armor value for PvE.
You seem to have forgotten Strength of Undeath and Spiteful Talisman in your calculations? Or am I just missing something?
I had Strength of the Undeath in an earlier version, but I removed it because I am mostly below the 50% threshold. However for now I added it just for greatsword.
Somehow I forgot to add Spiteful Talisman. Fixed this.
Introduction
Lately I have been theory crafting reaper builds to find an optimum between damage and survivability. The reason for this is that Reaper does not require berserker stats for its optimal damage and the rotation is easy.
(Sorry for the scientific report layout, but that makes it feel less like a wall of text.)
Setup
The four builds I calculated the DPS for are valkyrie, soldier, carrion and dire. All these builds use the exact same trait configuration. Spite for 25 stacks of might and damage boosters. Soul reaping for Dhuum fire and Reaper for 50% crit chance.
For these calculation I used:
- 2600 armor (too low for end game PvE?).
- 2 seconds per Gravedigger. 2.1 seconds for reaper auto attack (Including 15% speed boost).
- Close to death giving an average 10% direct damage boost.
- 100% shroud uptime.
- Fully self buffed is 25 might and 25 vul and target always burning. Also for Greatsword, even though this is not possible.
- Also included fury for the self buffed numbers.
- Flame legion runes and Balthazer runes in the combination they are optimal.
- Sigil of night, force, smoldering and bursting in the combination they are optimal.
- 20% condition Pizza.
Results
https://docs.google.com/spreadsheets/d/1CaiHdZF-YzC1yQbn9Y_XsgDc3-K_Hmp9nLG4T7ISbJM/edit?usp=sharing
Check spreadsheet for values after fixes
This shows that the highest DPS possible against 2600 armor enemies is 12.3k for greatsword and 10.7k for reaper shroud. Meaning that there is only a slight reason to spam greatsword 2 and not use the addition survivability for reaper shroud.
Moving on to Soldier (8.5k dps) vs Valkyrie (10.7k). There is a big boost for the addition ferocity (26%). This gap is even larger (30%) without party buffs . However due to Dhuum fire, this gap is smaller than expected.
Dire stats (9.0k dps) vs Carrior (9.9k dps) show Carrior is slightly stronger. But when taking into account the reduced Reaper uptime, then Dire will probably out damage Carrior.
Conclusion
So my conclusion is that reapers will cap their DPS at 12k average or 9k to 10k for much more tankier builds. So my question is how acceptable is this DPS compared to other classes?
Is it just me, or does Reaper allow for a good number of builds, but not any diversity in trait setup. Are there any reaper DPS builds that do not use Spite for solo play? The only reason it is included in the condition builds is to allow for 25 stacks of might (40% damage boost), which is the minor trait. All other traits from the Spite line are nearly useless for conditions.
Edit: Greatsword/dagger builds would not require Soul Reaping trait line.
Removed: – Including dungeon consumable (10% direct damage bonus).
Because this caused consumables to be applied twice.
(edited by Azzumy.7685)
Some constructive feedback:
In my opinion reaper is far from having overpowered damage. Greatsword is slow, this means that dodging will greatly reduce the DPS. On top of that if an enemy blocks or evades, then this skill goes on cooldown. If there is anything wrong then it is that the risk that this skill brings is rare in PvE and that the reward also feels not worth it.
Instead of nerfing it, it might also be possible to buff it. Make it deal 50% more damage when your life force is above 50% and make it drain 50% life force below 50% hp or 100% above 50% hp. This will solve the issue of spamming 2 all the time and will go nicely in the high risk for high reward category for both PvE and PvP.
If you really want to nerf something then a way to change this without affecting anything else is by inversing Close to death. Having 20% damage increase on targets above 50% health will both buff and nerf this attack.
My main was a guardian healer with nomad gear for a few months. When playing content with new players this was fun and useful, because I could compensate for their lack of dodge skill and lack of knowledge of how to heal back up. However, in parties with people that know the game better, all I did was giving buffs (stability) and watching them get resurrected within 2 seconds after dying (by burst or single hit).
During this time I noticed that dedicated healing was the most useless thing to focus on in guild wars. Parties require stability, which Druid does not have. Or enough DPS to rally someone, which Druid should lack. Or fast reviving, which can be done with runes of mercy.
My conclusion was that Berserkers are the best healers, because they can simply rally someone without losing any DPS. When there are no adds to rally, then sacrificing 10% damage for runes of mercy is the second best option. Any direct healing is only to compensate lack of skill.
Have you already tried this in practice? My main was a guardian healer for a long time. A guardian has 360 radius heal on dodge, healing breeze and mace symbol. With these skills I hit the players I want to heal most of the time.
Then I tried an elementalist, because in theory it has double the healing. However in practice I could barely hit any of my elementalist heals, because people dodge when they get damage and the 240 radius on water staff auto attack was too small.
In beta weekend 1 I noticed that revenant has some really strong healing, however the outgoing healing power did nothing. While waiting for beta weekend 2 I made full ascended gear for revenant, because I was sure it would be a lot better than guardian.
During beta weekend 2 the outgoing healing power finally worked and tried revenant in a dungeon. Here I noticed that people again dodge when they get damage (dodge=300 radius), so they kept dodging out of my tablet healing (240 radius). The advantage of revenant is that the skills have no cooldown, this allows some nice burst healing if they don’t dodge or when they walk back in after dodging. But when the boss did a large attack damaging my whole team, then I most likely had not enough energy to heal them. Resulting in a strong heal every 5 seconds with a high chance to miss.
Guardian is a lot better at healing, because it has constant healing from mace symbol+trait, this symbol has a very large range so most times it affects the whole party. When burst healing it needed it can dodge 3 times in a row (endurance on weapon swap) and repeat this quite often (vigor on crit). For the case where dodge is not possible (no endurance) I have healing breeze, which can even fully heal warriors.
My point is when theory crafting healing you have to decide the chance of applying it to someone (that is kiting). Resource management, revenant’s energy is used for everything, no energy no healing. Do you need communication with you team? If yes, then you most likely get instantly kicked for not having anything close to a meta build. If people do communicate they are probably good enough that they don’t need healing. The people that require healing (new players) have an automatic response to go kiting, independent on communicated tactics.
You probably do not want this informatie, but the best healing is to revive a downed player. Reviving out heals everything there is, except rally. So berzerker thief (stealth+self heal on crit) is the best healer in my opinion. Second to that is any meta berzerker build that kills a monster to rally their allies.
People will hate you if you go healing. Thiefs, elementalists and guardians will (at best) like you if you give them protection as long as it is not a light field. Everyone (except juggenaut) will love you if you give them stability or projectile block. People will also love you if you rally them, but they probably dislike you if you revive them directly.
Overall the best support would be a (non-ranger?) zerker. The second best would be a guardian greatsword with knight gear, mercy runes and 40% boon duration food. This gives a pull on GS, fast revives, high chance to aggro due to toughness, a very effective healing breeze, lots of stability, almost spammable quickness.
If after all this you find a working healing build (meta worthy) then please tell me, I love playing a healer.
Edit:
My final though about revenant is that it scales quite badly with healing power and has a very good base healing. So you can go berserker with increased outgoing healing power. This makes you 20-30% weakers than a zerker, but will allow you to heal relatively good. But even then protection, stability and rallying is way better.
(edited by Azzumy.7685)
I don’t use underwater weapons.
That is what I meant with 3 sigils. Currently I am using a signal of night and dungeon specific sigil in my main weapon set. And a sigil of bloodlust in my secondary weapon set.
It seems weird that you can use 3 sigils at a time.
When I place a sigil of bloodlust on my weapon, get some stacks and then swap weapon set. Than my sigil of bloodlust still gives bonus power.
As far as I can remember this behavior was different before the update and it was not in the patch notes. So I guess it is a bug.
Anyone noticed a 15% drop in both health and armor? (In AC)
Kinda funny how people only notice that their power is reduced.
The least Arena Net can do is say that they changed this.
I am only playing Ascalonian Catacombs with my guardian in Nomad gear. Before the update and after down scaling I had 531 healing power, around 1060 armor and 6k+ health.
After the update and after downscaling I have 381 healing power, 919 armor and 5,2k health.
All my items are fully repaired and as level 80 without down scaling I have the same stats as before.
Did they forget to place this in the patch notes or is it a bug? (What isn’t bugged after this update….)
I have the same issue. Sometimes the game freezes for a few seconds. This only happens like once a hour, but I died in dungeons a few times because of it.
OS: win 7
CPU: intel i5-4670k
GPU1: Geforce GTX 970 (sli)
GPU2: Geforce GTX 970 (sli)
Ram: 8GB