Guild Wars 2

So I made a Death Blossom condition build for fun, but a post in another thread the other day made me want to check the math on the amount of healing I’m doing before Assassin’s Reward is taken into account.

This is under perfect scenarios, with one major assumption that absolutely need someone to double check for me. All values taken from the GW2 wiki.

Assumption: Death Blossom takes exactly one second for entire animation.
(Someone please get me an actual value on this. I haven’t seen anyone anywhere note any after cast animation on most Thief attacks.)
Assumption: Player is running with +3 max initiative trait.

Signet of Malice

Death Blossom hits 3 times per second and costs 4 initiative.
In 15 seconds, Death Blossom can be activated 6 times.

This means Deathblossom can do 18 hits in 6 of those 15 seconds.
A single auto-attack chain completes 4 hits in 2 seconds.
When not using Deathblossom, auto-attack can do 16 hits during the remaining 9 of those 15 seconds.
That’s 34 hits in 15 seconds under perfect conditions.

Scaling for Signet of Malice is 0.05 per hit per point of healing power.

Settler’s Amluet and Jewel together give 644 healing power.
Cleric’s Amulet and Jewel together give 923 healing power.

644 * 0.05 = 32.2 health per hit
923 * 0.05 = 46.15 health per hit

133 health healed per hit with 0 healing power.
165 health healed per hit with 644 healing power.
179 health healed per hit with 923 healing power.

32 * 34 = 1088 extra health in 15 seconds
46 * 34 = 1564 extra health in 15 seconds

4522 health healed in 15 seconds with 0 healing power.
5610 health healed in 15 seconds with 644 healing power.
6086 health healed in 15 seconds with 923 healing power.

Average 301 health per second with 0 healing power.
Average 374 health per second with 644 healing power.
Average 405 health per second with 923 healing power.

Withdraw
Scaling is 0.6 per point of healing power.

4344 every 15 seconds with 0 healing power.
4730 every 15 seconds with 644 healing power.
4897 every 15 seconds with 923 healing power.

Average 289 health per second with 0 healing power.
Average 315 health per second with 644 healing power.
Average 326 health per second with 923 healing power.

For Withdraw to beat Signet of Malice
Withdraw > 32 hits with Signet of Malice at 0 healing power. (4344 > 4256)
Withdraw > 28 hits with Signet of Malice at 644 healing power. (4730 > 4620)
Withdraw > 27 hits with Signet of Malice at 923 healing power. (4897 > 4833)

I added Cleric’s out of curiosity, but I’d be willing to do a proper breakdown with Cleric’s for Pistol Whip and Unload if there is interest and numbers are provided.

And if you want to compare to Warrior healing abilities, go here: https://forum-en.gw2archive.eu/forum/professions/warrior/Warrior-Healing-Abilities-Hp-sec. But don’t forget this is without Assassin’s Reward taken into account.

Signet of Malice will perform far greater in group situations due to death blossom hitting up to 5 opponents 3x, and AoE fields like Caltrops and Choking Gas also tick SoM. In general if you’re running at least 500 healing power with this build, SoM will always outperform Withdraw.

Also Dagger Storm: blowing that in a zerg with SoM and some Toughness makes it real hard to die.

Although the SoM/Withdraw decision usually comes down to something other raw healing over time, in my opinion. Withdraw is the one to go to if you need the condition removal and/or you’re running something with an effect on-heal (like Vigorous Recovery, for instance), whereas SoM tends to be more strictly hit-based.

Viking Jorun.5413:

Signet of Malice will perform far greater in group situations due to death blossom hitting up to 5 opponents 3x, and AoE fields like Caltrops and Choking Gas also tick SoM. In general if you’re running at least 500 healing power with this build, SoM will always outperform Withdraw.

For now, I’m running this build in PvP with some success (although I don’t know how much of that success is due to the build or just my own twitch skills and game knowledge). But when a solo queue pops I advertise myself as a close point bunker. Often times that just means one person from the enemy team coming at me, maybe two.

Also, I’m trying to move away from using Caltrops because it’s not like I’m running a lot of Immobilize. One dodge roll and a few steps gets you out of Caltrops’ range. I’m currently trying out Scorpion Wire as it’s an extra CC for dealing with Stomps/Revives/Heals as well as keeping them in Caltrops, but I’m also considering dropping both of them and adding Signet of Agility along with Shadow Refuge for better group support for those times I’m in a team fight.

I’ve noticed that as you add more HP, SoM approaches 5% of health healed per DB (on a single target).

I am running over 1000 HP right now (apothecary and passiflora) and I am just about at 5% of health healed each time I use DB on a single target.

5 targets would mean 25% health for every DB.

The drawback I have with SoM, and I try very hard to make it work is not the amount of healing it gives – it’s good, and it seems to be balanced right.

The problem I have is that for that amount of healing, the disadvantage is that you have to be hitting something to get it.

There is so much CC in this game, that this becomes a major drawback, and I don’t think the healing amount was balanced taking this into consideration.

If you can’t hit, you can’t heal.

The difficulty in balancing a skill like this is that it’s going to partially be dependent on the players skill….I mean even more so than other skills.

A CC’d Thief using SoM heals = 0.

Additionally, the healing benefits from increased attack speed, which there is so little of in this game – Critical Haste doesn’t actually provide that much benefit (2 or 3 seconds every 30?) and to my knowledge, there is no other way to increase attack speed (there’s that crappy haste utility, but I never use it).

I think that SoM needs a 25-33% boost in healing per strike – at least for PvP. I do alright with it in some PvE, although dungeons are hard.

Proven.2854:

Viking Jorun.5413:

Signet of Malice will perform far greater in group situations due to death blossom hitting up to 5 opponents 3x, and AoE fields like Caltrops and Choking Gas also tick SoM. In general if you’re running at least 500 healing power with this build, SoM will always outperform Withdraw.

For now, I’m running this build in PvP with some success (although I don’t know how much of that success is due to the build or just my own twitch skills and game knowledge). But when a solo queue pops I advertise myself as a close point bunker. Often times that just means one person from the enemy team coming at me, maybe two.

Also, I’m trying to move away from using Caltrops because it’s not like I’m running a lot of Immobilize. One dodge roll and a few steps gets you out of Caltrops’ range. I’m currently trying out Scorpion Wire as it’s an extra CC for dealing with Stomps/Revives/Heals as well as keeping them in Caltrops, but I’m also considering dropping both of them and adding Signet of Agility along with Shadow Refuge for better group support for those times I’m in a team fight.

Caltrops is absolutely necessary for point bunkering, as it takes up the entire point. I used or run a similar build and act as mid point bunker while the other 4 on my team went close and far. It’s very effective at holding point.

Proven.2854:

Death Blossom hits 3 times per second and costs 4 initiative.
In 15 seconds, Death Blossom can be activated 6 times.

This means Deathblossom can do 18 hits in 6 of those 15 seconds.
A single auto-attack chain completes 4 hits in 2 seconds.
When not using Deathblossom, auto-attack can do 16 hits during the remaining 9 of those 15 seconds.
That’s 34 hits in 15 seconds under perfect conditions.

You can’t do both auto-attack and Deathblossom in the same 15s time frame, thus your “34 hits in 15 seconds” is inaccurate — even under perfect conditions.

It’s much easier to calculate than all that – just go hit something, and calculate the total percentage of health you get back when you land all 3 hits from DB on a single target.

I can tell you that as Healing Power increases, it approaches 5% of health returned per Death Blossom on a single target.

At just over 1000 Healing Power, I am practically right at 5% each DB I use (single target).

http://gw2skills.net/editor/?fYAQRAoY8YlcmyOHcy5E+5EyOjdSn4ZYhOlnCwqrA-jECBYiDimCMQk0hONtIas12FRjVLjIqWvEzcrIa1SB4suK-w

spike heal is always harder to use than heal over time.

like even if healing signet healed the same amount as healing surge over a period of time, healing signet will still be much much easier to use.

Sir Vincent III.1286:

Proven.2854:

Death Blossom hits 3 times per second and costs 4 initiative.
In 15 seconds, Death Blossom can be activated 6 times.

This means Deathblossom can do 18 hits in 6 of those 15 seconds.
A single auto-attack chain completes 4 hits in 2 seconds.
When not using Deathblossom, auto-attack can do 16 hits during the remaining 9 of those 15 seconds.
That’s 34 hits in 15 seconds under perfect conditions.

You can’t do both auto-attack and Deathblossom in the same 15s time frame, thus your “34 hits in 15 seconds” is inaccurate — even under perfect conditions.

You start with 13 initiative. You use 4 Deathblossoms. 4 seconds passed during that, so you got 4 initiative regenerated during that time. That’s a 5th Deathblossom.

Now, 5 seconds have passed, you only have one initiative, what do I do for the other 10 seconds?

This is why I asked for full cast animation times for Deathblossom. From my own informal testing, it came to 1 second. On the wiki they at least confirm that auto-attacking takes 2.07 seconds for a full chain, but even then I’m trusting that after cast was taken into account.

ipan.4356:

I’ve noticed that as you add more HP, SoM approaches 5% of health healed per DB (on a single target).

I am running over 1000 HP right now (apothecary and passiflora) and I am just about at 5% of health healed each time I use DB on a single target.

5 targets would mean 25% health for every DB.

The drawback I have with SoM, and I try very hard to make it work is not the amount of healing it gives – it’s good, and it seems to be balanced right.

The problem I have is that for that amount of healing, the disadvantage is that you have to be hitting something to get it.

There is so much CC in this game, that this becomes a major drawback, and I don’t think the healing amount was balanced taking this into consideration.

If you can’t hit, you can’t heal.

The difficulty in balancing a skill like this is that it’s going to partially be dependent on the players skill….I mean even more so than other skills.

A CC’d Thief using SoM heals = 0.

Additionally, the healing benefits from increased attack speed, which there is so little of in this game – Critical Haste doesn’t actually provide that much benefit (2 or 3 seconds every 30?) and to my knowledge, there is no other way to increase attack speed (there’s that crappy haste utility, but I never use it).

I think that SoM needs a 25-33% boost in healing per strike – at least for PvP. I do alright with it in some PvE, although dungeons are hard.

You need to dodge CC. When you dodge CC, you get 0 heals from SoM, but you also take zero damage from the CC and zero damage from the follow up attack. Under perfect conditions, you don’t have to dodge, so much of my post is theory crafting.

signet of malice should only be used for lots of people or withdraw wil always out heal it.

signet of malice uses best…..

caltrops
poison field(shortbow 4)
cluster bomb
daggerstorm
ambush
unload+ricochet

For a burst heal – I just don’t think it’s bursty enough.

It needs just a little more, like 2-3% more healing.

Part of it’s problem, is that to get it to work at all, you have to invest in full Healing Power – otherwise, it doesn’t work, and that means massive trade offs to dmg.

ipan.4356:

For a burst heal – I just don’t think it’s bursty enough.

It needs just a little more, like 2-3% more healing.

Part of it’s problem, is that to get it to work at all, you have to invest in full Healing Power – otherwise, it doesn’t work, and that means massive trade offs to dmg.

The thing is, SoM isn’t meant to be used at all. It’s meant to keep you actively engaged in combat by design, so using it is a counter to its design.

Viking Jorun.5413:

ipan.4356:

For a burst heal – I just don’t think it’s bursty enough.

It needs just a little more, like 2-3% more healing.

Part of it’s problem, is that to get it to work at all, you have to invest in full Healing Power – otherwise, it doesn’t work, and that means massive trade offs to dmg.

The thing is, SoM isn’t meant to be used at all. It’s meant to keep you actively engaged in combat by design, so using it is a counter to its design.

I’m not talking about activating the Signet – I’m talking about the amount it heals per strike.

It needs to be increased by a small amount.

There are, as I see it, two extreme types of calculations as far as theorycrafting goes: static theorycrafting and dynamic theorycrafting, with a whole bunch of stuff going in the middle. Static theorycrafting is based off of calculations made on in-game constants (i.e. the damage formula, base stats, etc) whereas dynamic theorycrafting is based entirely off of fluctuating values that will change from player to player and from fight to fight (i.e. the amount of hits you will get on opponents, the amount of damage you will take, skill of opponents, etc).

One of the largest criticisms that my theorycrafting gets is that it’s “too static”, that the game is extremely dynamic and such values are meaningless since they fluctuate too much. Unfortunately, that’s not really true; no matter how much direct damage your opponent does, increasing your toughness by, say, 100, will still decrease the amount of damage done to you by (for example) 2%. This is a constant, and the benefit of, for example, the marginal benefit from additional toughness is derived from the equation at the bottom of this post. Whether or not Power, coefficients, etc are constant doesn’t really matter- in any particular battle against anybody, the top and bottom always must remain the same; that’s the definition of the variable.

The difference between static and dynamic theorycrafting is the number of assumptions that you take. For instance, I must assume that P = P, which is justified by its own definition (otherwise there must be some grave flaw in the game’s coding). However, with dynamic theorycrafting, the more dynamic you get, the more assumptions that you have to make, more averages you have to use, the more possible statistical errors you have to consider, and so forth.

What I’m really trying to say here, is that the OP assumes that a lot of variables are constant or are constant enough for his/her purposes that he/she can make various assumptions that, realistically, can’t be made. If you do make those assumptions, then you’re either basing your ideas off of an unsupported personal bias, an extremely skewed statistical distribution, or off of variables that are overall fairly impertinent to the matter at hand when taken in the context of all of the other variables. That’s the real flaw in the OP, and that’s why his/her results are, for all practical purposes, completely meaningless.

Arganthium.5638:

What I’m really trying to say here, is that the OP assumes that a lot of variables are constant or are constant enough for his/her purposes that he/she can make various assumptions that, realistically, can’t be made. If you do make those assumptions, then you’re either basing your ideas off of an unsupported personal bias, an extremely skewed statistical distribution, or off of variables that are overall fairly impertinent to the matter at hand when taken in the context of all of the other variables. That’s the real flaw in the OP, and that’s why his/her results are, for all practical purposes, completely meaningless.

Then help me out here. What can I do to make it a better test? What should be cut?

I’m not saying either heal is better than the other. I want to know which is better. I do admit I shouldn’t have added the Cleric’s set in retrospect; It’d only be important for Unload. I also don’t have enough numbers on Thief animations and while I could time it myself I wouldn’t have the precision desired.

I also made sure I put in the last section about raw number of hits needed to outheal Withdraw. That alone doesn’t take into assumption any other variables; just the number attacks needed with the healing power gained from the two main healing power amulets and respective jewels slotted in. How to get to those thresholds, and whether or not they can be sustained, is another question I was hoping to move towards while using the fewest extra skills, like Caltrops.

I don’t use SoM, but I do use withdraw. If your math is correct, that means against a single target you’d have to hit 85% of the maximum possible hits to gain an advantage is health gain alone. Empirically that’s impossible (for me) due to various dodging/evading/blocking skills. In addition, sticking to a target that closely without any dodging from either side rarely happens. As I understand it, the condition ticks on the second also trigger SoM, so that might add an additional 15 hits in that time period if you haven’t accounted for that already, assuming you can keep at least 1 damaging condition on them for that entire period. From a non-health advantage, the curing of immobilize, cripple, and chill + the benefit from gaining an evade for .75 seconds + the “on heal” triggers far outweigh the benefits of SoM. In a group setting, SoM might pull ahead, but that is pretty conditional. I would much rather break immobilize in a 1vX setting as that will get you killed really quickly in that scenario.

For futher reference, autoattacking into a zerg with a short bow would yield about 45 hits once you include bounces in that 15 second period. With choking gas and cluster bomb you might get more, then again if you don’t get the full bounces you might get less.

It really comes down to preference. Do you like active heals more than passive heals and/or do you encounter groups of 2+ enemies that are closely packaged together more than single targets?

It is not the condition ticks, but the field pulses that trigger Signet of Malice and heal you. Works really nice with caltrops, just got more then 8k healing out of it with only the 300 Healing from Shadow Arts

frans.8092:

It is not the condition ticks, but the field pulses that trigger Signet of Malice and heal you.

Ah, well then it seems that unless there are a fair number of of foes willing to sit in your fields that these would have minimal additional benefit. Although if they will, it appears that there can be significant improvement.

Unless they’re forced to sit in a small area, it explains the video with the thief running SoM on a smokescreen-stealth-poison-caltrops build dotting a small group to death. Lot of fun watching that one.

Proven.2854:

Arganthium.5638:

What I’m really trying to say here, is that the OP assumes that a lot of variables are constant or are constant enough for his/her purposes that he/she can make various assumptions that, realistically, can’t be made. If you do make those assumptions, then you’re either basing your ideas off of an unsupported personal bias, an extremely skewed statistical distribution, or off of variables that are overall fairly impertinent to the matter at hand when taken in the context of all of the other variables. That’s the real flaw in the OP, and that’s why his/her results are, for all practical purposes, completely meaningless.

Then help me out here. What can I do to make it a better test? What should be cut?

I’m not saying either heal is better than the other. I want to know which is better. I do admit I shouldn’t have added the Cleric’s set in retrospect; It’d only be important for Unload. I also don’t have enough numbers on Thief animations and while I could time it myself I wouldn’t have the precision desired.

I also made sure I put in the last section about raw number of hits needed to outheal Withdraw. That alone doesn’t take into assumption any other variables; just the number attacks needed with the healing power gained from the two main healing power amulets and respective jewels slotted in. How to get to those thresholds, and whether or not they can be sustained, is another question I was hoping to move towards while using the fewest extra skills, like Caltrops.

Well, for starters, measure healing in HPS (healing per second). To do that with SoMalice, just multiply the amount of initiative you get per second on average by 3 (number of LDB hits), divide by four (initiative cost for LDB), and then divide by the cast time for LDB. Finally, multiply by 100 (health gained per hit from SoM). That should get you the HPS for SoMalice, unless there’s something that I’ve completely forgotten.

However, that’s not really all too accurate. See, in doing all of this, we’ve left out the number of hits that you’ll get on average with LDB anyways. Assuming that the previous number created gives you a number higher than that for Withdraw’s HPS (w/ casting time), then divide the HPS for Withdraw by the HPS for your previous SoMalice calculation. That’ll give you a good idea of the percentage of hits that you need to achieve the same healing per second of Withdraw, which is really the only variable you should care about in this entire calculation. If your number is, say, 90%, then you’re probably unlikely to get more HPS with SoMalice than you are with Withdraw. On the other hand, however, a value of 10% should be fairly easy to achieve, and so you should run SoMalice in that case.

As for slope with healing power, you can also measure that in HPS. Just multiply the healing coefficient for SoMalice by the average number of hits per second that you get with LDB and SoMalice, and that’ll give you your answer (I think). On the other hand, just divide Withdraw’s coefficient by 15 plus the cast time.