Guild Wars 2

Hi all

I did have some fun trying to compute a value of the sustain for a few new meta builds:

• chronobunker (before and after the last patch) (http://metabattle.com/wiki/Build:Chronomancer_-_Alacrity_Bunker)
• condi rev (http://metabattle.com/wiki/Build:Herald_-_Condition_Mallyx)

and for comparison, old meta builds:

• power warrior (http://metabattle.com/wiki/Build:Warrior_-_GS/Hammer)
• bunker guardian (http://metabattle.com/wiki/Build:Guardian_-_Bunker)

The sustain index 1 that I am computing is supposed to represent the amount of sustained damage the enemy team has to put up to overcome the sustained healing of the build. The sustained index 2 is trying to take into account armor and protection+damage reduction, so it represents the amount of power damage you need (while index 1 is valid for condition damage). I excluded all skills which would prevent point capping (therefore mimicking the build trying to fight on point). I also assume that the enemy has not unblockable/unavoidable skills. This is a very gross estimate, based on value/cooldown and not on a rotation, but it does give an idea.
Formula:
sustain index 1 = healing per second / (1 – evade/block uptime)
sustain index 2 = sustain index 1 * damage mitigation * armor/2200

Chrono before patch (using sentinel amulet)
I computed with and without perma-alacrity. While perma-alacrity is not easy to maintain, it is not too hard to have close to perma.

• healing per second: 543 without alacrity, 815 with alacrity
• evade/block uptime (includes dodges): 55% without alacrity, 77% with alacrity
• sustain index 1: 1200 without alacrity, 3564 with alacrity
• sustain index 2: 2461 without, 7311 with

Chrono after patch

• evade/block uptime (includes dodges): 43% without alacrity, 58% with alacrity
• sustain index 1: 947 without alacrity, 1923 with alacrity
• sustain index 2: 1944 without, 3947 with

I used sentinel amulet since this is the most often used one. However, this is not the one maximizing sustain. For the sake of comparison, chrono after patch with clerics has a sustain index 2 of 5474 with alacrity.

What we can notice immediately is how much alacrity adds to the sustain, by increasing the healing but mostly by increasing the invulnerability uptime. That is why I think the only thing left to nerf for chronobunker is their personal alacrity (either by tweaking the alacrity on shatter or the trait “improved alacrity”).

Condi rev

• healing per second: 764 (assuming no conditions when healing with mallyx heal)
• evade/block uptime (includes dodges): 71%
• sustain index 1: 2887
• sustain index 2: 4642

Now you immediately notice something interesting: condi rev has a higher sustain than bunker mesmer. And those condi rev values are without alacrity (since my scenario was defending a point solo). Also, condi rev cannot maintain his own perma-protection or perma-regen. This means that a condi rev supported by a mesmer will have an absolutely insane sustain, much higher than the mesmer himself. And it does this on a glass-cannon amulet.

For comparison, here are old meta builds:

Power warrior

• healing per second: 479 (optimistic, I assumed always 2 adrenaline bars)
• evade/block uptime (includes dodges): 28%
• sustain index 1: 668
• sustain index 2: 694

this is how you would expect offensive builds to work: they do high damage but die fast. Not a surprise that this build is not meta anymore. You would get higher values with the axe-shield variant, but probably not amazingly different.

Bunker guard

• healing per second: 1566 (this was really tough to compute since there are many sources of healing, I would not be surprised if I did a few mistakes)
• evade/block uptime (includes dodges): 24%
• sustain index 1: 2068
• sustain index 2: 3415

What should be the reference of bunker builds. As you see, chronobunker is still a bit above, mostly because of alacrity. Note however, that before asking for chronobunker to be nerfed, the overall power creep should be reduced too, since even bunker guard could not keep up with it.

So in conclusion:

• chronobunker has been nerfed significantly with the last patch, but still has a higher sustain than bunker guard, mostly thanks to the very high alacrity uptime
• condi rev still has the most ridiculous sustain. This kind of sustain should not be possible for a offensive build.
• power rev is even worse (higher invulnerability uptime), and is only “weaker” thanks to the amount of conditions in the current meta.

note: I did not include all details of the calculations because they are a lot of numbers. I do not claim my numbers are perfect, but I sincerely did my best.

Thanks for doing this! Could you post this on reddit too? So devs may notice.

nice job man!

do it for engi too

Silverkey.2078:

Hi all

I did have some fun trying to compute a value of the sustain for a few new meta builds:

• chronobunker (before and after the last patch) (http://metabattle.com/wiki/Build:Chronomancer_-_Alacrity_Bunker)
• condi rev (http://metabattle.com/wiki/Build:Herald_-_Condition_Mallyx)

and for comparison, old meta builds:

• power warrior (http://metabattle.com/wiki/Build:Warrior_-_GS/Hammer)
• bunker guardian (http://metabattle.com/wiki/Build:Guardian_-_Bunker)

The sustain index 1 that I am computing is supposed to represent the amount of sustained damage the enemy team has to put up to overcome the sustained healing of the build. The sustained index 2 is trying to take into account armor and protection+damage reduction, so it represents the amount of power damage you need (while index 1 is valid for condition damage). I excluded all skills which would prevent point capping (therefore mimicking the build trying to fight on point). I also assume that the enemy has not unblockable/unavoidable skills. This is a very gross estimate, based on value/cooldown and not on a rotation, but it does give an idea.
Formula:
sustain index 1 = healing per second / (1 – evade/block uptime)
sustain index 2 = sustain index 1 * damage mitigation * armor/2200

You should be able to come up with a ratio of condition damage to power damage a build is subjected to and use that to come up with a much better sustain index. You could do this by simply tracking end of match damage values in a excel spreadsheet, or the math way of taking popular meta comps and computing their condition damage and power damage outputs.

Silverkey.2078:

Hi all

I did have some fun trying to compute a value of the sustain for a few new meta builds:

• chronobunker (before and after the last patch) (http://metabattle.com/wiki/Build:Chronomancer_-_Alacrity_Bunker)
• condi rev (http://metabattle.com/wiki/Build:Herald_-_Condition_Mallyx)

and for comparison, old meta builds:

• power warrior (http://metabattle.com/wiki/Build:Warrior_-_GS/Hammer)
• bunker guardian (http://metabattle.com/wiki/Build:Guardian_-_Bunker)

The sustain index 1 that I am computing is supposed to represent the amount of sustained damage the enemy team has to put up to overcome the sustained healing of the build. The sustained index 2 is trying to take into account armor and protection+damage reduction, so it represents the amount of power damage you need (while index 1 is valid for condition damage). I excluded all skills which would prevent point capping (therefore mimicking the build trying to fight on point). I also assume that the enemy has not unblockable/unavoidable skills. This is a very gross estimate, based on value/cooldown and not on a rotation, but it does give an idea.
Formula:
sustain index 1 = healing per second / (1 – evade/block uptime)
sustain index 2 = sustain index 1 * damage mitigation * armor/2200

Chrono before patch (using sentinel amulet)
I computed with and without perma-alacrity. While perma-alacrity is not easy to maintain, it is not too hard to have close to perma.

• healing per second: 543 without alacrity, 815 with alacrity
• evade/block uptime (includes dodges): 55% without alacrity, 77% with alacrity
• sustain index 1: 1200 without alacrity, 3564 with alacrity
• sustain index 2: 2461 without, 7311 with

Chrono after patch

• evade/block uptime (includes dodges): 43% without alacrity, 58% with alacrity
• sustain index 1: 947 without alacrity, 1923 with alacrity
• sustain index 2: 1944 without, 3947 with

I used sentinel amulet since this is the most often used one. However, this is not the one maximizing sustain. For the sake of comparison, chrono after patch with clerics has a sustain index 2 of 5474 with alacrity.

What we can notice immediately is how much alacrity adds to the sustain, by increasing the healing but mostly by increasing the invulnerability uptime. That is why I think the only thing left to nerf for chronobunker is their personal alacrity (either by tweaking the alacrity on shatter or the trait “improved alacrity”).

Condi rev

• healing per second: 764 (assuming no conditions when healing with mallyx heal)
• evade/block uptime (includes dodges): 71%
• sustain index 1: 2887
• sustain index 2: 4642

Now you immediately notice something interesting: condi rev has a higher sustain than bunker mesmer. And those condi rev values are without alacrity (since my scenario was defending a point solo). Also, condi rev cannot maintain his own perma-protection or perma-regen. This means that a condi rev supported by a mesmer will have an absolutely insane sustain, much higher than the mesmer himself. And it does this on a glass-cannon amulet.

For comparison, here are old meta builds:

Power warrior

• healing per second: 479 (optimistic, I assumed always 2 adrenaline bars)
• evade/block uptime (includes dodges): 28%
• sustain index 1: 668
• sustain index 2: 694

this is how you would expect offensive builds to work: they do high damage but die fast. Not a surprise that this build is not meta anymore. You would get higher values with the axe-shield variant, but probably not amazingly different.

Bunker guard

• healing per second: 1566 (this was really tough to compute since there are many sources of healing, I would not be surprised if I did a few mistakes)
• evade/block uptime (includes dodges): 24%
• sustain index 1: 2068
• sustain index 2: 3415

What should be the reference of bunker builds. As you see, chronobunker is still a bit above, mostly because of alacrity. Note however, that before asking for chronobunker to be nerfed, the overall power creep should be reduced too, since even bunker guard could not keep up with it.

So in conclusion:

• chronobunker has been nerfed significantly with the last patch, but still has a higher sustain than bunker guard, mostly thanks to the very high alacrity uptime
• condi rev still has the most ridiculous sustain. This kind of sustain should not be possible for a offensive build.
• power rev is even worse (higher invulnerability uptime), and is only “weaker” thanks to the amount of conditions in the current meta.

note: I did not include all details of the calculations because they are a lot of numbers. I do not claim my numbers are perfect, but I sincerely did my best.

Extremely flawed argument. Especially when it comes to asking for nerfs. You have to account for skills. And class mechanics.

You can’t just craft a sustain value and base balance on that.

Also reapers and chrono didn’t exist during bunker guardian Era. Neither did the current spike damage

Rev was op due to resistance spam. Not the sustain. And I like how you kept auramancers out of your calculator

Hunter.4783:

Silverkey.2078:

Hi all

I did have some fun trying to compute a value of the sustain for a few new meta builds:

• chronobunker (before and after the last patch) (http://metabattle.com/wiki/Build:Chronomancer_-_Alacrity_Bunker)
• condi rev (http://metabattle.com/wiki/Build:Herald_-_Condition_Mallyx)

and for comparison, old meta builds:

• power warrior (http://metabattle.com/wiki/Build:Warrior_-_GS/Hammer)
• bunker guardian (http://metabattle.com/wiki/Build:Guardian_-_Bunker)

The sustain index 1 that I am computing is supposed to represent the amount of sustained damage the enemy team has to put up to overcome the sustained healing of the build. The sustained index 2 is trying to take into account armor and protection+damage reduction, so it represents the amount of power damage you need (while index 1 is valid for condition damage). I excluded all skills which would prevent point capping (therefore mimicking the build trying to fight on point). I also assume that the enemy has not unblockable/unavoidable skills. This is a very gross estimate, based on value/cooldown and not on a rotation, but it does give an idea.
Formula:
sustain index 1 = healing per second / (1 – evade/block uptime)
sustain index 2 = sustain index 1 * damage mitigation * armor/2200

Chrono before patch (using sentinel amulet)
I computed with and without perma-alacrity. While perma-alacrity is not easy to maintain, it is not too hard to have close to perma.

• healing per second: 543 without alacrity, 815 with alacrity
• evade/block uptime (includes dodges): 55% without alacrity, 77% with alacrity
• sustain index 1: 1200 without alacrity, 3564 with alacrity
• sustain index 2: 2461 without, 7311 with

Chrono after patch

• evade/block uptime (includes dodges): 43% without alacrity, 58% with alacrity
• sustain index 1: 947 without alacrity, 1923 with alacrity
• sustain index 2: 1944 without, 3947 with

I used sentinel amulet since this is the most often used one. However, this is not the one maximizing sustain. For the sake of comparison, chrono after patch with clerics has a sustain index 2 of 5474 with alacrity.

What we can notice immediately is how much alacrity adds to the sustain, by increasing the healing but mostly by increasing the invulnerability uptime. That is why I think the only thing left to nerf for chronobunker is their personal alacrity (either by tweaking the alacrity on shatter or the trait “improved alacrity”).

Condi rev

• healing per second: 764 (assuming no conditions when healing with mallyx heal)
• evade/block uptime (includes dodges): 71%
• sustain index 1: 2887
• sustain index 2: 4642

Now you immediately notice something interesting: condi rev has a higher sustain than bunker mesmer. And those condi rev values are without alacrity (since my scenario was defending a point solo). Also, condi rev cannot maintain his own perma-protection or perma-regen. This means that a condi rev supported by a mesmer will have an absolutely insane sustain, much higher than the mesmer himself. And it does this on a glass-cannon amulet.

For comparison, here are old meta builds:

Power warrior

• healing per second: 479 (optimistic, I assumed always 2 adrenaline bars)
• evade/block uptime (includes dodges): 28%
• sustain index 1: 668
• sustain index 2: 694

this is how you would expect offensive builds to work: they do high damage but die fast. Not a surprise that this build is not meta anymore. You would get higher values with the axe-shield variant, but probably not amazingly different.

Bunker guard

• healing per second: 1566 (this was really tough to compute since there are many sources of healing, I would not be surprised if I did a few mistakes)
• evade/block uptime (includes dodges): 24%
• sustain index 1: 2068
• sustain index 2: 3415

What should be the reference of bunker builds. As you see, chronobunker is still a bit above, mostly because of alacrity. Note however, that before asking for chronobunker to be nerfed, the overall power creep should be reduced too, since even bunker guard could not keep up with it.

So in conclusion:

• chronobunker has been nerfed significantly with the last patch, but still has a higher sustain than bunker guard, mostly thanks to the very high alacrity uptime
• condi rev still has the most ridiculous sustain. This kind of sustain should not be possible for a offensive build.
• power rev is even worse (higher invulnerability uptime), and is only “weaker” thanks to the amount of conditions in the current meta.

note: I did not include all details of the calculations because they are a lot of numbers. I do not claim my numbers are perfect, but I sincerely did my best.

Extremely flawed argument. Especially when it comes to asking for nerfs. You have to account for skills. And class mechanics.

You can’t just craft a sustain value and base balance on that.

Also reapers and chrono didn’t exist during bunker guardian Era. Neither did the current spike damage

Rev was op due to resistance spam. Not the sustain. And I like how you kept auramancers out of your calculator

This.

You can’t blindly try to calculate heals per second, mitigated damage per second and state that a class needs nerfing. While they do need nerfing, this is not the reason nor the direction it needs.

rev’s access to too much resistance, 85% stability uptime, and clones of some of the most key/best skills in the game (consume conditions, guard regen block, self alacrity/quickness, and resistance.. = necro/guard/mesmer/ele) makes it the lovechild of all the OP things other classes bring to the table. These need to be CAREFULLY finely tuned down to curb their effectiveness in PvP without nerfing it into oblivion.

mesmer is meta right now out of their sheer ability to maintain an easy to use rotation and simply sit on a cap. The game basically is, get your mesmers on cap, let them hold the status of the cap, your 2x rev and 1 ele with rotate and pressure out a player until people finally pressure the mes down or one of the revs fold. Usually the latter happens first.

I somewhat agree with curbing the personal alacrity, as that would fix a lot of the issue, but that would also heavily affect and almost nullify its viability in PvE, which is not a good idea.

I am impressed, you sir have done a great deed for the world.

A small note on bunker guard, as I am a high level bunker guard and have used it in many, many esls, bunker guard has great sustain but cannot hold its own against the amount of conditions and power burst currently available because of its VERY small health pool.

Also, would you mind computing and adding tempest as well? They are broken af bunkers too.

Sheobix.8796:

This.

You can’t blindly try to calculate heals per second, mitigated damage per second and state that a class needs nerfing. While they do need nerfing, this is not the reason nor the direction it needs.

Well you can, just not the way the OP did it. When I played swtor, the theorycrafters there had come up with a way to mathematically model the effects of heals and various defensive cooldowns on a combat situation, granted that game has combat logging (Anet pls) so they had more data to work with.
And now that I’m thinking aobut it, I’m gonna see if I can find a thread detailing how they did that.

Pls do it also for engi/scrapper. They have also ridiculous sustain.

I’d like to mention that 100% uptime of self alacrity is a feature of Chrono I wouldn’t like to see nerfed as it is the core aspect of the elite spec. Without damaging the team support I have a much better suggestion in my opinion:
Nerf the unalacritied cooldowns of the sustain skills until they are in line with the other classes. This means that it still has 100% personal uptime if builds don’t change and pushes out the same team support while lowering personal sustainability.

PS. Rev is balanced against condis hurting him during those invulns and not being able to attack for nearly all of that 71% uptime is also pretty poor, unless you go malix in which case invuln uptime goes way down.

+1

Good contribution.

It would be cool to see some similar math for Marauder Scrappers, Druids, and DS Elementalists, but I know those might not be as easy.

And maybe Reapers so we can all have a hardy laugh.

Arutha.9874:

I’d like to mention that 100% uptime of self alacrity is a feature of Chrono I wouldn’t like to see nerfed as it is the core aspect of the elite spec. Without damaging the team support I have a much better suggestion in my opinion:
Nerf the unalacritied cooldowns of the sustain skills until they are in line with the other classes. This means that it still has 100% personal uptime if builds don’t change and pushes out the same team support while lowering personal sustainability.

PS. Rev is balanced against condis hurting him during those invulns and not being able to attack for nearly all of that 71% uptime is also pretty poor, unless you go malix in which case invuln uptime goes way down.

He is talking about mallyx rev. Mallyx rev has more sustain than glint imo due to resistance.

NeXeD.3042:

Arutha.9874:

I’d like to mention that 100% uptime of self alacrity is a feature of Chrono I wouldn’t like to see nerfed as it is the core aspect of the elite spec. Without damaging the team support I have a much better suggestion in my opinion:
Nerf the unalacritied cooldowns of the sustain skills until they are in line with the other classes. This means that it still has 100% personal uptime if builds don’t change and pushes out the same team support while lowering personal sustainability.

PS. Rev is balanced against condis hurting him during those invulns and not being able to attack for nearly all of that 71% uptime is also pretty poor, unless you go malix in which case invuln uptime goes way down.

He is talking about mallyx rev. Mallyx rev has more sustain than glint imo due to resistance.

I am not sure he is as he said that he was in no way considering conditions in his calculations.
Sure we know that overall Mallyx is better but if you are just trying to maximise invuln and heals then Shiro and glint are probably best.

He literally links the mallyx build. As far as glint goes I’m not sure I would call that heal an invuln and it’s incredibly easy to counter.

Hunter.4783:

Silverkey.2078:

Hi all

I did have some fun trying to compute a value of the sustain for a few new meta builds:

• chronobunker (before and after the last patch) (http://metabattle.com/wiki/Build:Chronomancer_-_Alacrity_Bunker)
• condi rev (http://metabattle.com/wiki/Build:Herald_-_Condition_Mallyx)

and for comparison, old meta builds:

• power warrior (http://metabattle.com/wiki/Build:Warrior_-_GS/Hammer)
• bunker guardian (http://metabattle.com/wiki/Build:Guardian_-_Bunker)

The sustain index 1 that I am computing is supposed to represent the amount of sustained damage the enemy team has to put up to overcome the sustained healing of the build. The sustained index 2 is trying to take into account armor and protection+damage reduction, so it represents the amount of power damage you need (while index 1 is valid for condition damage). I excluded all skills which would prevent point capping (therefore mimicking the build trying to fight on point). I also assume that the enemy has not unblockable/unavoidable skills. This is a very gross estimate, based on value/cooldown and not on a rotation, but it does give an idea.
Formula:
sustain index 1 = healing per second / (1 – evade/block uptime)
sustain index 2 = sustain index 1 * damage mitigation * armor/2200

Chrono before patch (using sentinel amulet)
I computed with and without perma-alacrity. While perma-alacrity is not easy to maintain, it is not too hard to have close to perma.

• healing per second: 543 without alacrity, 815 with alacrity
• evade/block uptime (includes dodges): 55% without alacrity, 77% with alacrity
• sustain index 1: 1200 without alacrity, 3564 with alacrity
• sustain index 2: 2461 without, 7311 with

Chrono after patch

• evade/block uptime (includes dodges): 43% without alacrity, 58% with alacrity
• sustain index 1: 947 without alacrity, 1923 with alacrity
• sustain index 2: 1944 without, 3947 with

I used sentinel amulet since this is the most often used one. However, this is not the one maximizing sustain. For the sake of comparison, chrono after patch with clerics has a sustain index 2 of 5474 with alacrity.

What we can notice immediately is how much alacrity adds to the sustain, by increasing the healing but mostly by increasing the invulnerability uptime. That is why I think the only thing left to nerf for chronobunker is their personal alacrity (either by tweaking the alacrity on shatter or the trait “improved alacrity”).

Condi rev

• healing per second: 764 (assuming no conditions when healing with mallyx heal)
• evade/block uptime (includes dodges): 71%
• sustain index 1: 2887
• sustain index 2: 4642

Now you immediately notice something interesting: condi rev has a higher sustain than bunker mesmer. And those condi rev values are without alacrity (since my scenario was defending a point solo). Also, condi rev cannot maintain his own perma-protection or perma-regen. This means that a condi rev supported by a mesmer will have an absolutely insane sustain, much higher than the mesmer himself. And it does this on a glass-cannon amulet.

For comparison, here are old meta builds:

Power warrior

• healing per second: 479 (optimistic, I assumed always 2 adrenaline bars)
• evade/block uptime (includes dodges): 28%
• sustain index 1: 668
• sustain index 2: 694

this is how you would expect offensive builds to work: they do high damage but die fast. Not a surprise that this build is not meta anymore. You would get higher values with the axe-shield variant, but probably not amazingly different.

Bunker guard

• healing per second: 1566 (this was really tough to compute since there are many sources of healing, I would not be surprised if I did a few mistakes)
• evade/block uptime (includes dodges): 24%
• sustain index 1: 2068
• sustain index 2: 3415

What should be the reference of bunker builds. As you see, chronobunker is still a bit above, mostly because of alacrity. Note however, that before asking for chronobunker to be nerfed, the overall power creep should be reduced too, since even bunker guard could not keep up with it.

So in conclusion:

• chronobunker has been nerfed significantly with the last patch, but still has a higher sustain than bunker guard, mostly thanks to the very high alacrity uptime
• condi rev still has the most ridiculous sustain. This kind of sustain should not be possible for a offensive build.
• power rev is even worse (higher invulnerability uptime), and is only “weaker” thanks to the amount of conditions in the current meta.

note: I did not include all details of the calculations because they are a lot of numbers. I do not claim my numbers are perfect, but I sincerely did my best.

Extremely flawed argument. Especially when it comes to asking for nerfs. You have to account for skills. And class mechanics.

You can’t just craft a sustain value and base balance on that.

Also reapers and chrono didn’t exist during bunker guardian Era. Neither did the current spike damage

Rev was op due to resistance spam. Not the sustain. And I like how you kept auramancers out of your calculator

Well, he’s not calling for nerfs solely based on the sustain values (see his point about the overall power creep) and I don’t think anyone here says that you should. However, modelling of sustain like this is very helpful, even if it’s not perfect.

@ OP: Great post. Thanks for sharing and doing the maths.

I don’t think these simple maths are enough to discuss nerfs and balance, but they do give an idea. Then, we have also played games using those builds or against them, and I don’t think those results are really surprising for anyone.

I did not include tempest because the class is more rotation based so here, I would need to actually do rotations. I did not include scrapper because I never played the class so I have no clue of what does what (only class I haven’t played yet).

For the nerf to alacrity, I think the strength of the class in PvE is mostly linked to its outgoing alacrity/quickness more than its personal alacrity/quickness. The personal alacrity of course helps rotating faster and thus increase the outgoing one, but from the calculations I and other people in the mesmer forum did, the chronomancer contribution is currently very high and a nerf to personal alacrity will most likely not put the class in uselessness.

Finally, for the revenants, it is clear to me that the shield offers too much healing, and also that the revenant heals tends to be a tad too strong considering that there is 2 of them. Also, Riposting Shadows should not give a full dodge worth of endurance. I definitely think a-net should go slowly with nerfs, and that’s why I don’t like their idea of just one big patch and then 3 months or more of waiting.

For those who are interested:

Marauder scrapper
http://metabattle.com/wiki/Build:Scrapper_-_Marauder_Hammer

If you use your healing skills off cooldown and assuming you pick up one bunker down bandage every 4 seconds and around 50% protection uptime, so no protection from cc or teammates and no alacrity, you should get numbers like this:

sustain index 1: 1923
sustain index 2: 2731

and

Shatter Chrono
http://metabattle.com/wiki/Build:Chronomancer_-_Power_Shatter

Well this index, if flawed for bunkers already, is pretty much useless for squishy burst builds, since they simply can’t and shouldn’t stand on point at all, but just for the fun of it, according to this index shatter Mesmer has numbers like this:

these numbers are made under the assumption of 50% alacrity uptime, which I think is reasonable for a Mesmer without wells and of always having one clone up while healing

survival index 1: 697
survival index 2: 570

plus an amazing amount of condi clears equal to how often you can proc sigil of generosity

nicely done, but you forget that DPS got way higher as well.

Silverkey.2078:

The sustain index 1 that I am computing is supposed to represent the amount of sustained damage the enemy team has to put up to overcome the sustained healing of the build. The sustained index 2 is trying to take into account armor and protection+damage reduction, so it represents the amount of power damage you need (while index 1 is valid for condition damage). I excluded all skills which would prevent point capping (therefore mimicking the build trying to fight on point). I also assume that the enemy has not unblockable/unavoidable skills. This is a very gross estimate, based on value/cooldown and not on a rotation, but it does give an idea.
Formula:
sustain index 1 = healing per second / (1 – evade/block uptime)
sustain index 2 = sustain index 1 * damage mitigation * armor/2200

So I sat down and parsed out what you did. And well..

You’re doing it wrong.

You stated that your goal was to determine how much sustain damage (per second?) is required to kill a target.
The obvious answer to that question would be to use the TTK equation, but instead of doing that you removed all the variables from the TTK equation except for HPS and DTPS, and then tried to break the DTPS variable into opponent’s DPS and the class being measured mitigation (block/invuln)
Removing everything but enemy DPS, mitigation, and HPS means that the equation is only valid for a infinite length fight, and moreover it means that your equation will basically just spit out how much pre-mitigation DPS is required to equal the HPS Which isn’t even remotely practical because in a real match you only have so much time to clear a node.

A better approach would to just be to come up with a estimate of power and condi DPS for a typical teamfight, and run that against the various forms of mitigation to come up with a DTPS number and plug that into the TTK equation.

This formula is a nice black box model. It basically says, I don’t care what is inside, but I want the values at the ports. Typical Electrical engineering analysis.
While it may not include everything, it is a nice basis of what is going on in a fight.
Regardless of the TTK model, if HPS > DPS we can follow trends.

Interpret Interrupt.3824:

This formula is a nice black box model. It basically says, I don’t care what is inside, but I want the values at the ports. Typical Electrical engineering analysis.
While it may not include everything, it is a nice basis of what is going on in a fight.
Regardless of the TTK model, if HPS > DPS we can follow trends.

But I’m not sure that it’s even remotely correct. I can’t for the life of me understand what this index mean. I can understand from his description what he was trying to do, but I don’t think his formulas actually accomplish his goal.

If all you want is a basis, the simplest (yet accurate) method would be to just come up with a arbitrary ballpark estimate of DPS in a teamfight with a ballpark estimate of the ratio of condi to power, and simply use that in a standard TTK equation.

Crinn.7864:

Interpret Interrupt.3824:

This formula is a nice black box model. It basically says, I don’t care what is inside, but I want the values at the ports. Typical Electrical engineering analysis.
While it may not include everything, it is a nice basis of what is going on in a fight.
Regardless of the TTK model, if HPS > DPS we can follow trends.

But I’m not sure that it’s even remotely correct. I can’t for the life of me understand what this index mean. I can understand from his description what he was trying to do, but I don’t think his formulas actually accomplish his goal.

If all you want is a basis, the simplest (yet accurate) method would be to just come up with a arbitrary ballpark estimate of DPS in a teamfight with a ballpark estimate of the ratio of condi to power, and simply use that in a standard TTK equation.

I am no expert in computing stuff for video games, so I do not know the TTK equations. So I do not pretend my suggestion is better, it is just the only suggestion I had in mind to do this. Yes mine are based on an “infinite fight” idea, and do not take into account the number of HP, but I still think they are very intuitive.

The sustain index 1 is simply how much damage (assuming constant damage per second and after taking into account armor/mitigation for power damage or condition cleanse/resistance for condi damage) you need to deal to equate the healing of the build.

Out of this x damage, only a small percent will actually hit, for example only 42% if you fight a chronomancer. So the actual damage received per second is on average 0.42*x. When this damage equals the healing, you have
0.42x = HPS = 815 for chrono => x = 815/0.42 = 1941 (which comes to 1923 without the rounding errors).

So this is literally the minimum amount of damage you need to inflict before the enemy starts loosing HP “on average” (since obviously his life will be going up and down).

The second index is not as clear, but it is trying to also take into account damage mitigation and armor, so it represents the weapon strength*coefficient/second you need to deal assuming you have 2200 power and no crits. It is a less intuitive measure but still works well to compare different builds.

Silverkey.2078:

I am no expert in computing stuff for video games, so I do not know the TTK equations. So I do not pretend my suggestion is better, it is just the only suggestion I had in mind to do this. Yes mine are based on an “infinite fight” idea, and do not take into account the number of HP, but I still think they are very intuitive.

TTK is fairly simple. it’s TTK = HPS*MaxHealth / DTPS. Where DTPS is damage taken after mitigation.
I bring up MaxHealth because it’s hugely important.
Example: A hypothetical class could do 15k HPS with 70% block/invuln uptime, but if said class only has 1k health, it’ll die faster than a thief. Conversely a class with 30k health but no blocks/invuln and 0 HPS will take much longer to kill.

The sustain index 1 is simply how much damage (assuming constant damage per second and after taking into account armor/mitigation for power damage or condition cleanse/resistance for condi damage) you need to deal to equate the healing of the build.

Out of this x damage, only a small percent will actually hit, for example only 42% if you fight a chronomancer. So the actual damage received per second is on average 0.42*x. When this damage equals the healing, you have
0.42x = HPS = 815 for chrono => x = 815/0.42 = 1941 (which comes to 1923 without the rounding errors).

I gathered as much. So for the purpose of this, you’re simply assuming they are chaining blocks/invuln on CD?

The second index is not as clear, but it is trying to also take into account damage mitigation and armor, so it represents the weapon strength*coefficient/second you need to deal assuming you have 2200 power and no crits. It is a less intuitive measure but still works well to compare different builds.

coefficients vary from ability to ability and so do cast times. Just calc to the DPS and let people figure out if they can hit that number.

Silverkey.2078:

I am no expert in computing stuff for video games, so I do not know the TTK equations. So I do not pretend my suggestion is better, it is just the only suggestion I had in mind to do this. Yes mine are based on an “infinite fight” idea, and do not take into account the number of HP, but I still think they are very intuitive.

TTK is fairly simple. it’s TTK = HPS*MaxHealth / DTPS. Where DTPS is damage taken after mitigation.
I bring up MaxHealth because it’s hugely important.
Example: A hypothetical class could do 15k HPS with 70% block/invuln uptime, but if said class only has 1k health, it’ll die faster than a thief. Conversely a class with 30k health but no blocks/invuln and 0 HPS will take much longer to kill.

The sustain index 1 is simply how much damage (assuming constant damage per second and after taking into account armor/mitigation for power damage or condition cleanse/resistance for condi damage) you need to deal to equate the healing of the build.

Out of this x damage, only a small percent will actually hit, for example only 42% if you fight a chronomancer. So the actual damage received per second is on average 0.42*x. When this damage equals the healing, you have
0.42x = HPS = 815 for chrono => x = 815/0.42 = 1941 (which comes to 1923 without the rounding errors).

I gathered as much. So for the purpose of this, you’re simply assuming they are chaining blocks/invuln on CD?

The second index is not as clear, but it is trying to also take into account damage mitigation and armor, so it represents the weapon strength*coefficient/second you need to deal assuming you have 2200 power and no crits. It is a less intuitive measure but still works well to compare different builds.

coefficients vary from ability to ability and so do cast times. Just calc to the DPS and let people figure out if they can hit that number.

Silverkey.2078:

I am no expert in computing stuff for video games, so I do not know the TTK equations. So I do not pretend my suggestion is better, it is just the only suggestion I had in mind to do this. Yes mine are based on an “infinite fight” idea, and do not take into account the number of HP, but I still think they are very intuitive.

TTK is fairly simple. it’s TTK = HPS*MaxHealth / DTPS. Where DTPS is damage taken after mitigation.
I bring up MaxHealth because it’s hugely important.
Example: A hypothetical class could do 15k HPS with 70% block/invuln uptime, but if said class only has 1k health, it’ll die faster than a thief. Conversely a class with 30k health but no blocks/invuln and 0 HPS will take much longer to kill.

The sustain index 1 is simply how much damage (assuming constant damage per second and after taking into account armor/mitigation for power damage or condition cleanse/resistance for condi damage) you need to deal to equate the healing of the build.

Out of this x damage, only a small percent will actually hit, for example only 42% if you fight a chronomancer. So the actual damage received per second is on average 0.42*x. When this damage equals the healing, you have
0.42x = HPS = 815 for chrono => x = 815/0.42 = 1941 (which comes to 1923 without the rounding errors).

I gathered as much. So for the purpose of this, you’re simply assuming they are chaining blocks/invuln on CD?

The second index is not as clear, but it is trying to also take into account damage mitigation and armor, so it represents the weapon strength*coefficient/second you need to deal assuming you have 2200 power and no crits. It is a less intuitive measure but still works well to compare different builds.

coefficients vary from ability to ability and so do cast times. Just calc to the DPS and let people figure out if they can hit that number.

Crinn.7864:

Interpret Interrupt.3824:

This formula is a nice black box model. It basically says, I don’t care what is inside, but I want the values at the ports. Typical Electrical engineering analysis.
While it may not include everything, it is a nice basis of what is going on in a fight.
Regardless of the TTK model, if HPS > DPS we can follow trends.

But I’m not sure that it’s even remotely correct. I can’t for the life of me understand what this index mean. I can understand from his description what he was trying to do, but I don’t think his formulas actually accomplish his goal.

If all you want is a basis, the simplest (yet accurate) method would be to just come up with a arbitrary ballpark estimate of DPS in a teamfight with a ballpark estimate of the ratio of condi to power, and simply use that in a standard TTK equation.

Hi I believe TTK is time to kill? If so it forces a lot more assumptions on the calculator rather than what the OP seems to be trying to do.
Calculating HPS is defender’s class based and DPS is attacker’s class based so your returned numbers would be a 2d matrix which would then depend on number of builds… Much more confusing and depends on my build being exactly same as meta.

My understanding on this number is “the minimum dps you have to do to make the defender’s hp start to decrease.” We can ballpark our dps ourselves and if our dps > than this index we are winning attrition war.
So if given the chrono has 3947 for index 2, I understand that I need to put out more than 4000 dps, and if I can put out 5000 dps then I can drop him 1000 health/s. If i can sustain that over 15s that is guaranteed kill on 15k etc. and if I can burst it then less time. And again this does assume sustained fight not burst, but that is what will almost always happen in this bunker meta.