We all know that Mesmers do bad DPS, but I wanted to find out just how bad/good it is, and how different setups may affect that. I did a series of tests on the heavy golems on HotM, with no amulet, no trinkets and no runes. This is to prolong the encounter for as much as possible to smoothen out random variations due to critical hits etc. The build I used is the Fractals build, without Phantasmal Haste.
http://gw2skills.net/editor/?fhAQRAsc8al0npKtFpxVNcrNiqhgqukOQiIluTpBA
Each test was repeated 10 times and the amount of time it took to kill the golem was recorded. The results are presented below as the mean, standard error and standard deviation.
Sword AA Only:
As the title says, this was hitting the golem with only the sword AA. You may see this as a control.
Mean: 38.95 ± 0.57s; St. Dev.: 1.79s
Sword AA + Blurred Frenzy:
Blurred frenzy was used whenever it was off cooldown. This coincided nicely with 3 completed AA chains.
Mean: 36.28 ± 0.24s; St. Dev.: 0.75s
Blurred frenzy increases your DPS by a little, so you’ll want to use it if you don’t plan on dodging any big hits with it.
Sword AA + 1 Phantasmal Swordsman:
The clock was started after the swordsman was summoned, as it is common to summon it before reaching the boss and hitting it.
Mean: 23.37 ± 0.42s; St. Dev.: 1.31s
Unsurprisingly one phantasm hugely increases DPS.
Sword AA + 1 Phantasmal Swordsman + 1 Phantasmal Duelist:
The clock was started after the first phantasm was summoned. The duelist was chosen as it is the 2nd highest DPS phantasm after the swordsman.
Mean: 18.32 ± 0.48; St. Dev.: 1.52s
DPS increased again, but not as much as before, probably because of the extra time needed to summon the second phantasm. The Duelist also does less DPS.
Sword AA + 2 Phantasmal Swordsmen + 1 Phantasmal Duelist:
Mantra of Restoration was swapped out for the Signet, allowing 3 phantasms up. As before the clock was started after the first phantasm was summoned.
Mean: 17.18 ± 0.38s, St. Dev.: 1.20s
DPS hardly changed from 2 phantasms, which is not surprising since it takes an extra 3 seconds to summon the other 2 phantasms. So in short fights, 2 is enough. Comparing mean kill-times without the phantasm summoning time included is 17.32s for 2 phantasms and 14.18s for 3, so quite significant.
Greatsword AA:
This test was done at maximum range of the greatsword, which is our best ranged weapon.
Mean: 51.58 ± 0.48s; St. Dev.: 1.50s
The greatsword is significantly worse than the sword even at maximum range.
Mantra of Pain:
It’s interesting to see how using MoP affects our DPS with these two weapons. The clock was started with no charges on MoP, so the first 2.75s are spent charging it; this is to simulate an extended fight. Note that I am traited for Harmonious Mantras.
Sword AA + MoP Mean: 38.91 ± 0.43s, St. Dev.: 1.35s
Greatsword AA + MoP Mean: 43.14 ± 0.69s, St. Dev.: 2.18s
MoP hasn’t affected the DPS of the sword at all, so don’t bother using it. However it has hugely increased the DPS of the greatsword, so it’s a good idea to bring it for ranged encounters like, say, Tequatl.
To see how we compare to other classes, I also did the same DPS test on a thief. The build used was the D/D dungeon build on metabattle. However I removed the quickness trait as it has a huge impact on such short fight times, and substituted Revealed Training with a +175 Power rune as I didn’t have the trait. All attacks were done from behind, initiating with Cloak and Dagger, Backstab, 1 AA chain and restarting the cycle – ideal conditions for the thief.
Mean: 14.63 ± 0.21s; St. Dev.: 0.68s
The thief seems to do a lot more damage than the mesmer, being on par with having 3 phantasms up on extended fights, while in reality we probably only have 1 phantasm up on average. In this case a thief does about 70% more damage than us. However it’s notable that the 6/6/0/0/2 fractal build has substantially more survivability than the thief and offers excellent group support. Nonetheless it is near the maximum amount of damage we can do. It also isn’t always possible to attack from behind or stay in melee range.
I hope the information is useful and gives people an idea of what kind of damage we realistically do. The test isn’t perfect as removing the amulets would reduce our critical hit chance, hence reducing the effect of on-critical-hit traits. However, these are not very significant to begin with, for a Berserker build. I believe it is one step up from calculations.
Let me know if you have any comments on my methodology or spot any significant flaws. I may update this with other weapons or scenarios in the future, but I have no plans for now.
(edited by SPNKr.6208)