PristineChaos

I believe Chillblains was tested pre-3.1 when Blood used the Glyph of Blood Strike.
If I remember, tests were done on 10 Sarth and there was a significant increase in Heart Strike damage (back when HS was affected by the glyph).

 

Xentik

I've just done some FS Weapon-speed dependence calculations and am getting some pretty interesting numbers. Specifically, if we are only trying to maximize the DPRP of FS, it is in your best interest to take the slowest weapon possible, even to the point of favoring much lower DPS weapons.

The calculation I did was as follows (including all formulas for clarity):

Definitions
FS(Base)=(0.55 * [WDPS + AP/14*2.4] + 138)*(1-0.5*__isoffhand?__) + SotVH(FC))*
(BlackIce)*(MercCombat)*(GlacierRot)*(BoTN)*(TunderStalker)*(1+NoCS*__isoffhand?__)
FS(DW,NonCrit) = FS(Base,MH)+FS(Base,OH)
FS(DW,Crit) = [FS(Base,MH)+FS(Base,OH)]*(2+GoG)*(CritMeta)
FS(DW,Mean) = FS(DW,NonCrit) * (1-Crit) + FS(DW,Crit)*(Crit)

Calculation
AP = 6000
Crit = 25%
MHDPS = 178.8, MHSPD = 1.5
Spec = The same spec I used in my earlier posts for Frost FSvsDC

FS(DW,Mean)(MH DPS, MH SPD) == FS(DW,Mean)(MH DPS - deltaDPS, MH SPD+deltaSPD)

Solve for deltaDPS as a function of DeltaSPD, then choose 1.1 for deltaSPD (which gives a 2.6 speed weapon). Using the function I get that a 103 dps 2.6 spd weapon is equivalent to a 178.8 dps 1.5spd weapon when it comes to frost strike.

I'd appreciate it if someone could sanity check my numbers and see if this is real or just a stupid error on my part.

 

Bloody_sorcerer

Totally makes sense, since the average weapon damage of the two is roughly the same, and that's the part that matters
103 * 2.6 = 267.8 damage
178.8 * 1.5 = 268.2 damage
Since FS only cares about average weapon damage, the two are roughly equal.

 

Xentik

Of course, I can't believe I forgot that. Thanks for sanity checking me.

Since this behaviour should also hold true for Obliterate and Blood Strike, I'd think that this would essentially throw Fast OH's out the window for a ToT build, and that you might even be better off digging an old, slower OH instead of your shiny new fast one. (e.g. Razorscale Talon > Remorse by ~200 per strike, ~6 per RP which is roughly a 5-6% increase in D and DPRP).

I think I'll take a look at BS and Ob's numbers to see if this theory holds water.

[Note: The 5-6% increase is MH only, so changing OH from Remorse->Razorscale Talon would actually garner a 2-3% increase in FS DPS]

 

Sakuratei

While going for slow weapons in that manner is better fir the strike itself, Weapon dps affects our overall dps a lot more to compensate it. It's not like this concept is new, all melee classes factor in weapon speed when they calculate the dps with a weapon, but for all classes the weapon dps outweights the speed in nearly all cases.

 

OriginalMemnock

So far on the PTR, here are my preliminary results on the DW testing I've done. First I did this allw ith my transferred toon... here is my armory so you can see what I'm currently wearing:

Yagyu

Spec = 10/54/7

All the testing I've done on the server is with the Ebon Hold boss dummy and no buffs (no horn and never used unbreakable armor). All test were approximately 10 minutes each. Currently on live my top DPS on said dummy is around 3400 dps in Frost and Unholy.

Now on to my DW test. The 1h weapons I used to test are not the best ever, but they are what I have available. They are:

Slow /Fast:

MH - Torment of Banished
OH - Bloodcrush Cudgel

Fast/Fast:

MH - Bloodcrush Cudgel
OH - Stoneguard (I know it's a tanking weapon but it's the only other fast 1h I have)

Enchants were MH Fallen Crusader, OH, Cinderglacier or Razorice

With Slow/Fast using Fallen Crusader and Cinderglacier enchant combination I averaged a whopping 3900 dps. That's with a level 200 slow hand folks!

With Slow/Fast using Fallen Crusader and Razorice I averaged 3500 dps.

With Fast/Fast using FC and Cinder I averaged 3200 dps.

With Fast/Fast using FC and Razor I averaged 2900 dps.

It's clear to me that a Slow/Fast combination with Fallen Crusader MH and Cinderglacier OH is the best combination.

Now as far as glyphs, I used the following different combinations for the tests:

Frost Strike + Oblit + Icy Touch
Frost Strike + Oblit + Howling Blast
Frost Strike + Oblit + Disease

Using HB to refresh frost fever didn't flow quite as smooth as I had hoped. Another shocking surprise is how well a rotation glyph of Disease worked. It took me a while to get a good flow with that combination, but once I had the timing down for hitting Pestilence at the right time to refresh diseases, I got very nice results. Using Icy Touch glyph I got very nice results as well.. and I guess the difference is just keeping a good rhythm of refreshing diseases.

The results I got even surpassed the new blood build on the PTR so I expect the nerf bat to hit before this goes live, but it's CLEAR the DW is BACK!

 

Sylari

Are you sure of this? I'm having difficulty getting my mods in game to differentiate between offhand and mainhand strikes, but swapping between [Torment of the Banished] and [Gladius] ( using a [Hailstorm] in the mainhand at all times ) in the offhand only yields an increase of about 19~% in the offhand with the epic, despite Torment having double the tooltip damage ( 265 average damage with the gladius , 545 average damage with the Torment ), having stats, and being slower than the gladius.

Also interesting is my hailstorm's average hit was 566 , but my ToT offhand strikes were never more than about 65% of the mainhands damage.

Even with Gladius in the mainhand and torment in the offhand my offhand strikes never exceeded 70% of the mainhand strike's damage.



edit:


Dummy tests aren't really the best, but I noticed you never tried slow/slow or FC/FC, any reason why, seems the build you're trying doesn't have anything that would make a fast offhand desirable at all.

 

AtheistGod

The slower aspect of torment actually throws off your tooltip. Because AP scaling is the same with both The actual value used for strike with gladius is higher than what your tooltip has while with torment it is lower. The difference is the 150 average damage difference before multipliers. The fact that you get such a significant increase (19%) shows that it is using the offhand. The stats on a one hand weapon is no where near enough to cause such a difference.

 

Sylari

Ahh, I forgot about that, my mistake.

And yes, slow seems to be very dominant in the offhand.. my tests are by no means definitive or even that effective as my mods don't seem to be functioning as well as they should, but messing around with it a bit further and keeping the gladius in the mainhand and swapping between Torment and Hailstorm showed about a 5% difference in oblit damage.

 

Nizari

Ugh, that's disheartening... I really liked this spec. I'll have to play around and see what works and what doesn't. Even though I did post almost the same build and information a page earlier, the latest information about Chill Blaines not working is useful.

I know they changed "Torment the weak" to attack speed reductions as well, has anyone simply tested the glyph with Frost Fever (without chilblains) on a target?

http://sigs.guildlaunch.net/sig.php/419258FqvWj.png

 

teiglin

That's not correct. Multiple haste buffs that grant a percentage stack multiplicatively, including windfury/IIT, the 5% from the IIT talent, heroism/bloodlust, and the sum your haste rating on gear/procs. So if you have 20% haste on gear, IIT, and are in unholy presence, your total haste is 1.2*1.2*1.05*1.15 or 73.88%. With just IIT, your 100 hits become 126 hits, not 125. It's pretty easy to see from the swing speed listed on your character sheet, though I haven't done any large sample of tests to verify that the character sheet is correct.

 

kidpreacher

I have done some testing on the Frost server, and with Slow/Slow FC/FC weapons I noticed two things.

1. Unholy Blight trained does not work with Death Coil dot. It's broken untrained it applies the dot of 0 damage trained it does not. Not sure why it just didn't work for me.

2. FC does not stack I did numerous tests and it didn't stack at all. Now I had a pretty close to constant up time, but when it overlapped it simply refreshed the previous FC.

However FC/FC slow/Slow was about 200-300 dps less than RI/FC. Cinderglacier was so far behind the other two I quit after 3 minutes of testing.

Slow/Fast I found to be about 300-400 dps behind the Slow/Slow. I will post my spreadsheet later, but this is a very rough calculation no buffs.

Talent Calculator - World of Warcraft was the spec and glyphs I used that produced the highest total dps for me. Unholy presence was also used, and produced a significant dps increase.

Perma Ghoul Frost Strike was okay, but a tougher rotation to fill.

EDIT: Howling Blast Glyphed is basically a wasted glyph now as you will have to decide between it and OB, and the only time you would choose HB over OB is on multiple mobs. If the glyph was somehow changed to allow it to attach Blood Plague as well, than Howling Blast would be worth it, but it seems that both HB/DRW are the 2 51 point talents you will rarely see specced into.

My rotation on the above was
IT/PS/OB/BS/BS/FS
OB/OB/BS/FS until diseases need to be reapplied.
- then I ran into a problem of one more FS or toss refresh my diseases.

 

OriginalMemnock

I didn't try slow/slow simply because of a lack of a decent slow weapon. I had to use a level 200 weapon as is for my slow weapon. In regards to dual FC, if I'm not mistaken, FC has an internal cooldown which would make dual FC moot. I could be wrong though, but that was always my impression.

 

Naivedo

The reason I use Frost Strike over Obliterate for Deathchill is because of these following talents Rime, Subversion, Annihilation, and Dark Conviction. Of course, some of these talents also increase Frost Strike's critical strike chance but no where as close to Obliterates. Just Rime and Subversion alone is +24% more critical strike with Obliterate over Frost Strike. Deathchill is best with the ability that has the lowest critical strike and Obliterate has at least a 24% higher chance of being wasted with the right talents.

 

OriginalMemnock

Okay so I got a hold of a second slow weapon and did some slow/slow tests using the same criteria I used before, and quite honestly, I did not notice any significant difference in using slow/fast versus slow/slow.

I also tried unholy and blood presence and blood came out on top unquestionably.

 

Edimasta

Are you sure about that? Anyone else has seen this curious occurance with the Offhand damage?
This could mean that, when GC told that OH dmg would be "roughly cut in half" that it can never be more like 65% of the MH damage... would be interesting to know.

And you all keep telling that Dummy tests are not the best... but everyone keeps doing it.
So what do you suggest? Or what are you all waiting for?

Lets go and raid a bit... ^^

 

Kyruski

Dummy tests can be fine for finding out the mechanics of a spell/ability, but when it comes to finding out the highest DPS spec, it is in no way the best idea. Also, we are trying to halt all spec discussion until we figure out all the mechanics/everything in the "Need to test" section of the OP. Once we complete that list, finding a spec is a lot easier and a little more mathematical without there being a lot of speculation and "My spec does xxxxDPS".

 

Sylari

Yes I'm sure, I must have stood perpendicular to the dummy and hit him with OB/BS for ten minutes ( not the best test, but meh ) and the highest offhand strike was 69.874% of the mainhand's damage, even when using a low level grey mainhand and a ilvl 200 offhand.

I'm going to try testing a bit more tomorrow, need to find a slower/higher level epic and a faster/lower level mainhand to try and stretch the gap as much as possible.

 

AtheistGod

It's because of other gear. That equalizes them. Try with nothing on and a high end weapon in the offhand and a low main hand. they are much closer.

 

Whatevr

I did some more testing just now. This test was to figure out whether it's better to use BS for death runes vs oblit. I didn't even have to dps for 2 mins to find out the answer for this.

These are the two specs I used and yes I know "Combat dummy is stupid testing" but when comparing two specs with each other that both involve mostly the same talents I think it's safe to say it's fine to test it this way.

Both specs have no glyphs affecting any abilities.

Weapons used were 2x [Razorscale Talon] main hand FC offhand RI (RI was up 100% of the time)

[Sigil of Awareness] was used.

I never used unbreakable armor in any of the tests.

Howling blast wasn't ever used during this test.

HoW was up 100% of the time and Arcane Torrent was used every time I could use it. Deathchill was used primarily on FS. Sometimes I'd get unlucky and I'd get back to back KM procs so it would get used on IT instead.

I had to use the ebon knight's training dummy due to the fact that too many dks were using ebon plague on the boss dummy.

Spec Used: http://talent.mmo-champion.com/?deat...version=10026#

Test 1
Using Oblit for death runes

IT -> PS -> oblit -> BSx2 -> RP dump
oblit x3 -> RP dump


yes some dumbass decided to come to my dummy out of all the ones in there that were empty and start attacking it and putting up ebon plague. I had to stop the test early (only about 300k short but it looks pretty clear which one wins)

Test 2
Using the same spec I decided to try and use BS for Death runes to see how it goes. Here are the results.



IT -> PS -> Oblit -> BS x2 -> RP dump
Oblitx2 -> BS x 2 -> RP dump

I was wearing 2pc t8 and 2pc t7 for this test. I get the feeling that even if I had lost the 2pc t7 bonus oblit would still be on top due to the large amount of crit it gains from the frost tree too.

I know somebody did some math with BS being more damage than oblit however I don't think they factored in the crit bonus that oblit gains compared to BS. Hopefully this can help out if needed. If you need anymore info let me know.

edit: Changed original post to now show only using 1 spec and using BS for the 2 death runes instead of oblit.

 

Kyruski

Comparing DPS isn't the way to go with Comparing BS vs. OB/Anything Else with Death Rune Usage, Especially with different specs and such a low test time. You want to compare the damage per rune of the strike/ability used. I did some Calcs earlier in the thread with BS vs. OB here and BS vs. IT here . The was a problem with my tests and yours(unless they've changed it by now). That problem is BS is still doing 50% per disease instead of the 25% per disease (unless they've changed this already). It turned out that IT > BS per rune under the circumstances that I had, but the problem with both BS and IT for Death Runes was fitting them all into one rotation via latency (most likely PTR latency). If on Live we are able to fit the OB>OB>IT>IT and extra FS into one rotation, then that will surely Outweigh OB>OB>OB or OB>OB>BS>BS.

 

Whatevr

Yea the ptr still has BS as 50% per disease so it's already doing more than it should be doing for when it goes live. On top of that I agree that using 2 death runes in the rotation as single rune abilities mixed in with a FS in there will cause problems in the rotation. Between the GCD and latency it will delay the rotation a little bit. However it may be able to work if the extra RP generated from that last sequence was drained after the new diseases are up. It would look like this:

IT -> PS -> OB -> BSx 2 -> RP dump (HB if rime procced, this is what i typically did)
OB x2 -> IT x 2 -> FS x1 (remaining RP is dumped after the repeat of the first cycle)

I typically had a bit of time between the first and second cycle and often found myself not having enough rp to fill in until the runes were ready (i also didn't have the FS glyph so that could be one reason). That's just my thought though.

 

kidpreacher

I am not sure where this thought comes up. Enhance shamans use Dual Slow because WF and SS are based on weapon damage not weapon dps. Mutilate rogues use Slow/Slow because Mutilate is based on Weapon damage not weapon dps. Combat rogues and Fury Warriors also use Slow/Fast or Slow/Slow on the basis that most of their attacks are based on weapon damage not weapon dps. If you saw any of the above classes with Fast/Fast, you would tell them to learn their class. Top End Weapon damage tops out the dps charts on Windfury/SS/Hemo/Mutilate I don't think I need to continue here.

Therefore since a majority of our dps is coming from attacks on Cooldown based on Weapon Damage and not weapon dps it should be pretty easy to figure out that the higher weapon damage is the choice to go ignoring dps. Perhaps I am misunderstanding his statement, but this seems to be a commons misconception viewed out there with many melee users.

The mechanics of FS/OB/BS are all top end weapon damage the higher the weapon damage the more Damage per swing giving us a higher dps total output. With Icy Talons/Unholy presence Slow/Slow becomes even better.

Whatevr, I did the same test with a different spec, ignoring Epidemic and using unholy presence and Blood Presence and came out with similar numbers to your with Unholy Presence producing about 200 more dps. I don't like wasting points on Plague strike to get to epidemic. Epidemic allows a longer Rune Dump, but in my tests it turned out less of a dps output. Granted Blood Presence without epidemic was a pain to get a cycle in, it forced me into Unholy were the rotation became easier. I ran the exact same rotation.

 

Kyruski

Actually, Mut rogues in PvE use Slow/Fast for the Poison Procs (and a few others). Combat rogues do the same.

Also with Icy Talons and Unholy Presence, They favor neither Fast/Fast, Slow/Fast, Slow/Slow, or Fast/Slow since it increases the speed of the melee swing by a flat percentage.

 

Sakuratei

While this is true, the quote I posted discussed about speed in the Offhand. Our strikes hit for roughly 0.575 of a main strike, and strikes in general hit for ~60% weapon damage + fixed value + % of total damage of before mentioned factors regarding to diseases.

I don't think that 0.1 in speed in regards to weapon damage will outweigh the weapon's acutal dps when looking at the DK's dps as a whole and not just strikes (Seeing as how preliminary tests show white attacks as high as 25% of damage done), but I've never put myself into the maths and may as well be wrong.