Guild Wars 2

Spent few hours recording each phantasm and then using video editing software to record the time between attacks with Phantasmal Haste and without. Then, in HotM (no armor, no trinkets, no sigils, no power traits) against a dummy golem (open field, not against a wall) I recorded 20 non-crits from each phantasms and took an average for each. Finally, I divided the average damage over the time between attacks. Here are attack rates:

Attack rate (untraited):

• Swordsman – 4.5 sec
• Berserker – 7.1 sec
• Duelist – 7.5 sec
• Warlock – 6.8 sec
• Warden – 13.7 sec
• Mariner – 10.3 sec
• Mage – 7.7 sec
• Whaler – 9.7 sec
• Disenchanter – 4.4 sec
• Defender – 4.6 sec

Attack rate (traited):

• Swordsman – 4.5 sec
• Berserker – 7.1 sec
• Duelist – 6.5 sec
• Warlock – 5.7 sec
• Warden – 12.2 sec
• Mariner – 10.3 sec
• Mage – 6.5 sec
• Whaler – 8.3 sec
• Disenchanter – 3.8 sec
• Defender – 4.0 sec

Attack rate (difference):

• Swordsman – 0% faster
• Berserker – 0% faster
• Duelist – 15% faster
• Warlock – 19% faster
• Warden – 12% faster
• Mariner – 0% faster
• Mage – 18% faster
• Whaler – 17% faster
• Disenchanter – 16% faster
• Defender – 15% faster

I added a second table on wiki for Phantasmal Haste beside the current one where I listed the times between attacks. The time is the amount of time between 1st damage hit of the phantasm. In case of swordsman, the time between every hit. In case of duelist or any multi-hit phantasm, the time between the very 1st hit for each unload.

What a surprise, duelist and swordsman are the highest DPS.

But iZerker is good guys, I swear.

/rolls eyes

Duelists dps is sooo much higher…
/roll eyes

God forbid there are situations with more than one enemy. Never seen that before.

colesy, you are a bit like schroedinger’s cat… Good luck in your box. ^^

In my tests, berserker always connected 3 out of 4 possible hits in an open field on a non-moving golem. If a golem was moving in the same direction as the berserker’s spin then all 4 hits connected. Also, those possible 12 stacks of bleed on everyone around the Warden and that possible 120% more DPS on the Warlock from conditions… mmmm

frifox.5283:

Also, those possible 12 stacks of bleed on everyone around the Warden and that possible 120% more DPS on the Warlock from conditions… mmmm

10%

Xyonon.3987:

frifox.5283:

Also, those possible 12 stacks of bleed on everyone around the Warden and that possible 120% more DPS on the Warlock from conditions… mmmm

10%

x12 ^^

Huh, looks like duelist falls behind if you don’t take Phantasmal Haste (but, of course, EVERY build should take that, right?).

But iZerker is horrible, guys, I swear.

(Note to all the rational people out there – I realize the difference between duelist and zerker looks to be within the margin for error. I just don’t care for blanket statements from folks who reply without reading.)

Kaiyanwan.8521:

Xyonon.3987:

frifox.5283:

Also, those possible 12 stacks of bleed on everyone around the Warden and that possible 120% more DPS on the Warlock from conditions… mmmm

10%

x12 ^^

Ah i tought he talked about the wardens stacks for the warlock xD my bad ^^’’

http://wiki.guildwars2.com/wiki/Condition

Count the unique conditions…

An edit full of wisdom up there. ^^

What? Has Swordman being fixed?

Kaiyanwan.8521:

http://wiki.guildwars2.com/wiki/Condition

Count the unique conditions…

An edit full of wisdom up there. ^^

Ye I was too slow >.< or you too fast

Btt: Everything like I expected to be. Swordsman and Warlock by far the highest single target DpS, Pistol is just meh, Zerker best for AoE, Warden medicore bot AoE and reflection. The others are 0 damage 100% condition / utility. Perfect

colesy.8490:

What a surprise, duelist and swordsman are the highest DPS.

But iZerker is good guys, I swear.

/rolls eyes

what are you trying to say here? isword/iduelist are both single target when izerker is aoe AND applies cripple….

He wants to point out that the iZerker has about the same single target dps as the iDuelist with an incredible possibility of an average of 555dps with multiple targets with this setup.

That’s it I think. It would be insane if he wanted to point out the damage advance of the iDuelist vs the iZerker (wait, next he will come up with bleeds…)

So Warlock (with Phantasmal Haste) hitting a target with conditions would do:

5 conditions – (97 base x 50%) + 97 = 145.5 dps?
6 conditions = 155 dps
7 conditions = 165 dps

At 7 conditions, is this our highest DPS phantasm?

Oh wait…didn’t take into account bleeds from duelist and iswordsman.

So flat dps numbers are pretty much useless. What you collected that is useful is the interval between the first hits, and you can use that to calculate the real dps.

I recently finished compiling a list of all mesmer skills and their coefficients, I’ll post the phantasm numbers here:

• Phantasmal Swordsman: 7
• Phantasmal Berserker (Per Hit, up to 4): 2.03
• Phantasmal Warlock (0 Conditions, *1.1 per): 5.4
• Phantasmal Duelist(Per Hit): 1
• Phantasmal Warden (Per Hit): 1.1
• Phantasmal Mage: 2.9

Using those numbers you can calculate true dps. The one issue I have with your numbers, frifox, is that you got different numbers for swordsman traited and untraited. I’m fairly certain that there is 0 change for traiting the swordsman, you may want to check your videos again.

As a note for anyone wondering what these numbers mean.

Damage = Weapon Damage * Power * Skill Coefficient / Target Armor

Those numbers are the specific skill coefficients for each phantasm.

One more note: The damage variation you are experiencing with the swordsmen is most likely due to inconsistencies with attack periods and your overall recording interval. That is why straight dps numbers are always 100% useless. The only thing that matters is interval between attacks.

So the actual hit/sec increased by PHaste is:

• Phantasmal Berserker – 5.1%
• Phantasmal Defender – 11.1%
• Phantasmal Disenchanter – 15.5%
• Phantasmal Duelist – 13.3%
• Phantasmal Mage – 15.5%
• Phantasmal Swordsman – 8.8%
• Phantasmal Warden – 9.9%
• Phantasmal Warlock – 18.5%
• Illusionary Mariner – 0%
• Illusionary Whaler – 15.4%

And so many players thought the trait would do less than told.

But the actual DPS increase is:

• Berserker – 6.3%
• Defender – 14.2%
• Disenchanter – -24%
• Duelist – 13.2%
• Mage – 16.6%
• Swordsman – 7.5%
• Warden – 9.1%
• Warlock – 18.5%

How comes, that there’s a difference ?
pDisenchanter attacks (15%) faster but has (24%) less DPS ?

As for the order:

1. Swordsman – 158 dps
   Single target, close range
2. Duelist – 113 dps
   Single target, ranged, multihit, good combo
3. Berserker – 111 dps
   AoE, close range, some hits, bad combo, cripple
4. Warlock – 97 dps
   Single target, ranged, with avg + 30% damage, it’d be ranked #2
5. Warden – 87 dps
   AoE, melee range, multihit, bad combo, anti-projectile, bugged immobile

Seems to be a fair list.

Thanks for testing.

[ Edit ]
^ @Pyro
You saying Warden has a coeff of 1.1 per hit per target? A total coeff of 13.2 (no death) which’d outrank any other ?

Is there bleeding, reflection, AoE and/ or cooldown included?

@Nretep: My numbers simply say how much damage each hit does. Reflection, AoE, cooldown, bleeding, and attack speed are not included in my numbers. They are purely skill coefficients.

In the example of the warden, that’s a cumulative skill coefficient of 13.2 spaced out over a significant period of time with a significant cooldown.

It would be nice to capture the numbers for a single attack “round” on each phantasm as many don’t live long enough for a second round of attacks (due to AoE, mob death, or shatters).

In this case DPS is not that useful since the damage is not sustained. However damage for the attack chain is useful. Example: if Duelist unloads for 2k and Swordsman for 1.5k, then duelist is better for the scenario where the phantasm dies after the attack. However skill c/d comes into play now as you can only repeat the single attack per the c/d.

I wonder how one calculates the DpS :o

Xyonon.3987:

I wonder how one calculates the DpS :o

I suspect you’re trying to be funny — but you did touch on the root of what I’m discussing.

If you assume the phantasm dies after it’s initial attack chain (i.e. doesn’t get a second round of attacks), then the DPS equation will use the skill cooldown (for the phantasm) in the denominator. The DPS calculations above use the phantasm’s actual skill recharge in the denominator.

I understand math is hard — let me know if you need me to clarify it some more

juno.1840:

It would be nice to capture the numbers for a single attack “round” on each phantasm as many don’t live long enough for a second round of attacks (due to AoE, mob death, or shatters).

That’s what Pyro just said/ did above your post.

@Pyro
so it’s just the coeff of a single attack of the respective phants. Well, their recharge isn’t in, but with those values and the OP’s new PHaste timings, anyone should be able to calc the values as wanted. Since you listed “per hit”, adding bleeding is possible as well.

Still interesting numbers. pWarden has the strongest attack cycle with AoE and anti-projectile attack. And it’s attack doesn’t even end then the target dies (it completes its cycle). No wonder I like this one

Xyonon.3987:

I wonder how one calculates the DpS :o

Well, pWarden deals total of 1200 damage and initiates its attack every 10s = 120 d/s.

Thx Nretep — so using Pyro’s coefficient numbers (this time for the full attack chain):

Phantasmal Swordsman: 7
Phantasmal Berserker (4 hits): 8.12
Phantasmal Warlock (3 Conditions): 8.7
Phantasmal Duelist(8 hits): 8
Phantasmal Warden (12 hits): 13.2

Cooldowns for each weapon (no 20% wep traits as that’s just a raw increase of 20%):

Phantasmal Swordsman: 20s
Phantasmal Berserker (4 hits): 20s
Phantasmal Warlock (3 Conditions): 18s
Phantasmal Duelist(8 hits): 20s
Phantasmal Warden (12 hits): 25s

Taking the coefficients and dividing by the cooldown gives this picture:

Phantasmal Swordsman: 0.35
Phantasmal Berserker (4 hits): 0.406
Phantasmal Warlock (3 Conditions): 0.483
Phantasmal Duelist(8 hits): 0.4
Phantasmal Warden (12 hits): 0.528

Now multiply by the weapon damage to get DPS (for this scenario of one attack chain and then phantasm is dead — assumes 2000 power and 2600 armor which is reasonable — it’s just scaling):

Phantasmal Swordsman (1000 dmg): 269
Phantasmal Berserker (1100 dmg): 344
Phantasmal Warlock (1100 dmg): 408
Phantasmal Duelist(1000 dmg): 308
Phantasmal Warden (900 dmg): 365

Unfortunately these numbers don’t match well with the numbers already calculated so either there’s a different assumption for numbers like power and armor — or the coefficients are wrong — or ????

@juno

…or you’re comparing apples with oranges. The numbers calculated in the op were sorta random numbers that don’t mean anything over a long period of time. You just calculated some specific things assuming only 1 attack from the phantasm.

Pyroatheist.9031:

@juno

…or you’re comparing apples with oranges. The numbers calculated in the op were sorta random numbers that don’t mean anything over a long period of time. You just calculated some specific things assuming only 1 attack from the phantasm.

A sanity check on my calculations:

iZerker = 344 DPS (when it completes a single attack chain and dies)

Total damage for the attack chain = 344 DPS * 20s = 6880

That’s reasonable for the numbers I see when I launch my iZerker. However my zerker crits often and hard — so maybe the coefficients are a tad high? How’d you calculate them?

Just wanted to follow up on the pistol vs gs.
8 bleeds does add up.
Running around with 6 might (sigil) and 25 illusions gets bonus bleeding to 524 per volley. Having a group with banner of strength/25 might gets that to 859 more bleed dmg.

IFF traited the pistol can also do
burning: 5.2k
poison: 1.7k
life steal: 1.6k dmg
confusion: 2.6k dmg per tick
The other combo fields aren’t dmg oriented.
So around 900 to 3k to 10k (confusion vs alpha) bonus damage.

That’s some significant damage. Of course it is single target but that was part of cosleys argument? Most dungeons are skip trash kill single target. Of course the pzerker is aoe but the gs 1 and 2 are basically singlet target vs the sword cleave. Personally, I like to melee but I understand why ppl would want to use the GS at times. The pistol is significantly better at single target than the power numbers would show.

Wow, someone actually understood my point? It’s as if there are actually people on this forum who don’t just go “lalala you’re insane I play how I want”.

I’m sick of trying to get through to bads.

nobody denies that some phantasms are better against golems in the myst…..
If target is attacking, things change.

Why not starting reconrding with lot of data how long phants survives to deal their dps?

Also aoes but we already know that.

colesy.8490:

Wow, someone actually understood my point? It’s as if there are actually people on this forum who don’t just go “lalala you’re insane I play how I want”.

I’m sick of trying to get through to bads.

Then don’t. You’re the one starting the issue, again, in this thread and you wonder why people troll you ? Just let it drop.

We get it, if you want to be optimal on DPS you need iDuelist and iSword, but if you want to AoE, iZerker is better. Furthermore, iZerker damage is not that bad, BUT yes (oh god…!), iDuelist and iSword is better for DPS. Oh, also, Sword MH is the best for damage. Did I get it right ? Do you feel happy ?

We get your point. That’s all, case closed.

juno.1840:

Pyroatheist.9031:

@juno

…or you’re comparing apples with oranges. The numbers calculated in the op were sorta random numbers that don’t mean anything over a long period of time. You just calculated some specific things assuming only 1 attack from the phantasm.

A sanity check on my calculations:

iZerker = 344 DPS (when it completes a single attack chain and dies)

Total damage for the attack chain = 344 DPS * 20s = 6880

That’s reasonable for the numbers I see when I launch my iZerker. However my zerker crits often and hard — so maybe the coefficients are a tad high? How’d you calculate them?

Honestly Juno, I have no idea why you’re calculating based on a period of 20s. Just calculate damage per single attack of the phantasm, then use the attack period of the phantasm to calculate dps.

By calculating this random number over 20 seconds, you’re just making things complicated.

2.03*1000*2300/2800= 1667.5 per hit, non crit. Assuming 90% crit power and 3 hits landing, all crits, you get 12006 damage. This does seem a bit high honestly, so I’ll go recheck my damage coefficients when I have time later today.

Since my english is not the best: coefficients are the ratios? :o Like 500 base damage +0.15xpower?

Btw the question I made for “how do you calculate dps” proved once again that many ppl can’t do it. Just wanted to see who’s actually able to do math right. Sry if it looked like a trollpost, it wasn’t ^^

Another question about the “ratios” / damage. Is it just me or are some phantasms damage nubers completly wrong at the description? E.g. iWarden deals 4k while iZerk does 370 … uhm … whut?
Also do they scale the same as all the other skills? I noticed after checking all sword (and GS after sword) skills that power increases the damage of every skill by the same % amount.

Anyone knows more?

@Xyonon: Read my posts in this thread. Pay particular attention to the text in italics, and you should learn what the damage coefficient means.

Pyroatheist.9031:

juno.1840:

Pyroatheist.9031:

@juno

…or you’re comparing apples with oranges. The numbers calculated in the op were sorta random numbers that don’t mean anything over a long period of time. You just calculated some specific things assuming only 1 attack from the phantasm.

A sanity check on my calculations:

iZerker = 344 DPS (when it completes a single attack chain and dies)

Total damage for the attack chain = 344 DPS * 20s = 6880

That’s reasonable for the numbers I see when I launch my iZerker. However my zerker crits often and hard — so maybe the coefficients are a tad high? How’d you calculate them?

Honestly Juno, I have no idea why you’re calculating based on a period of 20s. Just calculate damage per single attack of the phantasm, then use the attack period of the phantasm to calculate dps.

By calculating this random number over 20 seconds, you’re just making things complicated.

2.03*2*1000*2300/2800= 1667.5 per hit, non crit. Assuming 90% crit power and 3 hits landing, all crits, you get 12006 damage. This does seem a bit high honestly, so I’ll go recheck my damage coefficients when I have time later today.

The 20s is not random — it’s the recharge of the phantasm skill itself (i.e. every 20s you can summon an iDuelist).

My math was for a very specific use-case: Phantasm dies shortly after summoning, getting off a single attack-chain. It’s a valid use case in some fights where phantasms die to AoE or by direct attack from a mob.

The recharge time of the skill is important — example that the iWarden is 25% less frequent than an iDuelist. It effects the ability to crank them out, and so raw damage for a single attack chain occurs less frequently.

It was mostly an exercise to satisfy my curiosity and provide more data — not to prove any particular point. I’m quite pleased that most of the phantasms are in the same arena with regards to damage output.

DuckDuckBOOM.4097:

Just wanted to follow up on the pistol vs gs.
8 bleeds does add up.
Running around with 6 might (sigil) and 25 illusions gets bonus bleeding to 524 per volley. Having a group with banner of strength/25 might gets that to 859 more bleed dmg.

IFF traited the pistol can also do
burning: 5.2k
poison: 1.7k
life steal: 1.6k dmg
confusion: 2.6k dmg per tick
The other combo fields aren’t dmg oriented.
So around 900 to 3k to 10k (confusion vs alpha) bonus damage.

That’s some significant damage. Of course it is single target but that was part of cosleys argument? Most dungeons are skip trash kill single target. Of course the pzerker is aoe but the gs 1 and 2 are basically singlet target vs the sword cleave. Personally, I like to melee but I understand why ppl would want to use the GS at times. The pistol is significantly better at single target than the power numbers would show.

Key words “if traited”… and it appears you also made assumptions about available combo fields. Lastly bleed damage doesn’t do much when the target is already capped at a full 25 stacks — and requires Sharper Images from Dueling line (great trait but I imagine some builds don’t use it?).

We could all theory-craft a scenario where any phantasm out-DPS’s another (except maybe for iMage — not sure how ANet let that one through QA).

The iDuelist is a great phantasm especially under the scenario you described above. However if you start to say it’s “mandatory” then you imply by association that a very specific gear and trait setup are also “mandatory”. Makes you wonder why ANet even made traits and gear itemization to begin with? (yeah that was sarcasm)

IFF stands for If and only if. I know that not everyone has it traited.

I understand that it doesn’t work out that way all the time. Ethereal are pretty common from pretty common mesmer skills though. I didn’t assume everyone has the others available. I included all possible combo fields that can do dmg and usually at least 1 shows up in a group of 5.

Unlike Colsey I have never stated that GS users are bad and you should never use it. However, the pistol has the potential to be a lot better than the raw numbers from above and I agree with colsey that it is our best dps phantasm. The only other phantasm like that is the warlock. The rest are basically as is. That shouldn’t stop you from playing the way you enjoy but it should be some facts to think about when and IF you ever decide to try and max your damage.

Hope you guy don’t misunderstand me. The numbers in op are BASE damage dealt by phantasms on average. Almost every phantasm can be better than the other in one situation or another. The whole purpose of this study was to give me a good idea as to how hard phantasms hit at minimum then use this knowledge and adapt to the encounter/situation to take advantage of other sources of DPS even better. This is how mesmers are, we are NOT warriors where all our damage comes from direct hits.

Pyroatheist.9031:

The one issue I have with your numbers, frifox, is that you got different numbers for swordsman traited and untraited. I’m fairly certain that there is 0 change for traiting the swordsman, you may want to check your videos again.

You are correct. I recorded 10x attacks and untraited Swordsman attacks between 4.200 to 5.033 sec and traited attacks between 4.050 to 4.917 sec, which leads me to believe that timings I observed could be within range of possible attack time but not represent the actual average attack rate.

I will redo the timings and update OP/Wiki later today with the range or average times between attacks once I’m sure they’re accurate.

@frifox: You use the term ‘base damage’. This does not exist. There is no such thing as ‘base damage’. Please refer to my earlier post for the damage formula.

I really appreciate all the research!

@juno just one point on Sharper Images – it’s a minor trait at 15 points, so it’s not optional. Everyone who goes 15 points into Dueling has it.

The minimum damage that trait can do is 42 per second for five seconds for each crit. So when my duelist shoots 8 times, I usually get 6 or 7 bleeding stacks at my crit chance. So at a minimum it’s about 250 additional dps for the duration of the bleed.

Adding a bit of might and 20% condition duration can turn this from 1250 damage to closer to 2000 over the course of 6 seconds. So that’s a pretty big impact on dps even without traiting or using combo fields.

In theory, the warden could get 12 crits, but players don’t stand around long enough for that. And the 3 or 4 crits that the berserker can get are actually a reasonable contribution to overall dps.

eldenbri.1059:

I really appreciate all the research!

@juno just one point on Sharper Images – it’s a minor trait at 15 points, so it’s not optional. Everyone who goes 15 points into Dueling has it.

The minimum damage that trait can do is 42 per second for five seconds for each crit. So when my duelist shoots 8 times, I usually get 6 or 7 bleeding stacks at my crit chance. So at a minimum it’s about 250 additional dps for the duration of the bleed.

Adding a bit of might and 20% condition duration can turn this from 1250 damage to closer to 2000 over the course of 6 seconds. So that’s a pretty big impact on dps even without traiting or using combo fields.

In theory, the warden could get 12 crits, but players don’t stand around long enough for that. And the 3 or 4 crits that the berserker can get are actually a reasonable contribution to overall dps.

It’s optional in that you’d have to put 15 points into Dueling — I always have at least 15 in dueling as I love dodge clones and sharper images. I was just stating that there may be some builds that do not use the dueling line at all.

For PvE, the warden can definitely apply a lot of bleeding — more so than the duelist. Totally agree that is NOT the case in WvW or PvP.

My other point on bleed stacks is they have no effect when your opponent already has 25 bleeds — very common for world bosses in PvE. However in WvW the bleeding is definitely a great DPS add.

I love Duelist in WvW, especially when fighting stealthy thieves as a couple duelists will kick kitten when he pops.

Just as a heads up folks, consider my damage coefficients for the phantasms as incorrect. It seems there needs to be more testing done before I can actually figure out the damage calculation for phantasms.

Ok, after some initial tests… iZerker also doesn’t benefit from Phantasmal Haste
I’ll redo the test from scratch to be 100% sure, because if this is true… gg anet

You can’t compare phantasms by a single number. Any of them has its advantages and disadvantages. Different situations require different weapons.
If you prefer another weapon than another mesmer, don’t critize him/her, he’ll be more suited in different situations.

Furthermore, do not mix PvE and PvP/ WvW.

I appreciate all testings and results, but can’t stand the posts saying “that is better than this because …” and “if you include this, it makes that better”.
We can discuss about underpowered phantasms/ skills/ traits, but don’t compare working ones.

P.S.: Mirror Images is better than Feedback.

Ok, after few more hours in HotM I figured out the timings for most phantasms. Tomorrow I will finish it up and add the timings for the disenchanter/defender and underwater phantasms. For now, wiki contains swordsman/zerker/duelist/warlock/warden/mage timings. OP is updated as well.

As far as Phantasmal Haste,

• Swordsman – 0% faster
• Berserker – 0% faster
• Duelist – 15% faster
• Warlock – 19% faster
• Warden – 12% faster
• Mage – 18% faster

Also, something interesting I’ve noticed. In LA during prime time some phantasms will attack +/- 0.5 seconds from the average yet some will attack at the same interval no matter what. In HotM, the timings are much more stable. During peak times they varied at most +/- 0.1 sec. During night time, the intervals were within 0.01 sec of each other (!). From this I can only assume that some of our phantasms can “lag” if the current server load is too high.

frifox.5283:

As far as Phantasmal Haste,

• Swordsman – 0% faster
• Berserker – 0% faster
• Duelist – 15% faster
• Warlock – 19% faster
• Warden – 12% faster
• Mage – 18% faster

excuse me but arent you slowly getting near old data about PH that is posted forgot if in bug section or wiki?

Just to save you some work in case you didn t know.

Old data on wiki listed the time between the end of phantasms attack animation and start of the next one. That doesn’t reflect the actual attack rate, just the time phantasm spends idling between their attack chains. What I was researching the the rate at which phantasms attack. Also, I believe the old data may not be entirely accurate as seen with the iZerker. I can only assume that the original tester ran into the same problem I did earlier, where some phantasms attack at same intervals and some vary up to +/- 0.5s on every attack. If you use only one sample data and not average of of 10 – 20 samples then your observed time has a chance of being up to 1s off.

Anyways, both Wiki and OP are now updated with data on all phantasms. Everything recorded should be accurate up to 16ms (60 fps).

frifox.5283:

Old data on wiki listed the time between the end of phantasms attack animation and start of the next one. That doesn’t reflect the actual attack rate, just the time phantasm spends idling between their attack chains. What I was researching the the rate at which phantasms attack. Also, I believe the old data may not be entirely accurate as seen with the iZerker. I can only assume that the original tester ran into the same problem I did earlier, where some phantasms attack at same intervals and some vary up to +/- 0.5s on every attack. If you use only one sample data and not average of of 10 – 20 samples then your observed time has a chance of being up to 1s off.

Anyways, both Wiki and OP are now updated with data on all phantasms. Everything recorded should be accurate up to 16ms (60 fps).

The old data certainly was accurate when I made it, and what it shows is the recharge time, since recharge is from the end of the skill to the beginning of the next one, and that is the actual number that the phantasmal haste trait changes. Things may have changed now, but I did get that data by counting frames in adobe premier.

I also advise you to completely remove your dps numbers from the OP as they are meaningless. What you should put there instead is the attack rate numbers.

Yep, I’m aware of that. I was more interested in how much faster the attack rates were, not just the recharges. Edited OP too.

frifox.5283:

Attack rate (untraited):

• Swordsman – kitten

Attack rate (traited):

• Swordsman – kitten

this forum filter….

so izerker is the second Best phantasm for dps.

And if my point is true on the fact zerker survives more than swordman….
Not to mention it adds cripple.

Well GS seems not so bad after all.

Almost all phantasm can be best dps if you use them in the correct situation. For example, iWarden ranks 5th on my DPS list yet there are many cases where it will hit harder than any phantasm out there: http://www.youtube.com/watch?v=dvtsmUM5DIo

After doing this research on phantasms I came to a conclusion, which I probably knew already but didn’t really pay much attention to, that it’s pointless theorycrafting or ranking phantasm DPS. It’s almost like comparing apples to oranges if you try to.

^
that’s why pyro said to remove your DPS ranking. It’ll make players like colesy say that pSword is better than pWarden.