5.25.2006

More on Sword Spec Modeling

There has been some discussion and validation of Sword Specialization mechanics on the Rogue forum recently, and the results are exactly the same as I stated in both 3/2/06 and 3/3/06 post.
  1. Sword Spec proc off only swords.
  2. OH sword spec proc triggers an instant extra MH attack.
  3. Special attacks triggers an instant extra MH attack.
  4. Sword spec proc is an normal mainhand attack and will cancel attack-in-motion and resets your MH attack timer.
  5. Sword spec will NOT proc off enchant/weapon/poison procs.
There are 3 cases that will trigger a sword spec proc: 1. MH normal attack 2. OH normal attack 3. Special attack. And it will be the determining factor in modeling sword spec proc dps.

Case 1: MH normal attacks

MH attack triggers another instant normal attack from sword spec. No swing timer is "wasted" as the sword spec proc is instant. Thus you gain 100% of the damage.

Effective MH sword spec proc damage = 100% * MH damage.

Case 2: OH normal attacks

OH attack triggers an instant MH normal attack.

t=0: Attack initialized. Both MH and OH swing at the same time.
t=OH_Speed: OH attacks. Procs sword spec. MH attacks instantly and resets the swing timer.

Thus, swing timers will reset to initial engagement state everytime OH attack procs sword spec, with the wasted MH swing time equal to OH speed.

Effective OH sword spec proc damage = [1 - (OH Speed/MH Speed)] * MH damage

Yes, your OH sword spec proc will do less effective damage, as the only "time"/"damage" you gain is the part you get for free but not yet swung. i.e. MH speed minus OH speed.

Case 3: Special attacks

Similar to OH proc modeling, assuming you aint doing anything stupid like SnD/Evis/Rupture.

Effective MH sword spec proc damage = [1 - (MH Speed/4)] * MH damage

Example:

2.8 Speed MH, 1.8 Speed OH.

MH Sword Spec Proc = 100% of MH damage
OH sword spec proc = 36% of MH damage
Special sword spec proc = 30% of MH damage

The overall sword spec proc effective damage (weighted average) will be 54%.

23 comments:

Anonymous said...

Hey! :)
That's an awesome work! Grats!
So after reading all this we could officially say that Fist Spec>Sword Spec?
And what do you think about the following builds regarding dmg output?

1)Sword(MH)+Dagger(OH) with sword+dagger combat build?

2)Fist(MH)+Sword(OH) with fist+sword combat build?

Thanks! :)

VitaminC said...

Well, it really depends, but intuitively, I believes sword spec > fist spec due to 1) additional procs 2) white damage scales much more nicely.

I don't think spec'ing two different weapons will help your dps by a whole lot, especially when you take into acount the scarcity of the available weapons.

I would say the general weapon choice rule for spec'ing heavy in combat tree would be 1) dw sword spec dw swords if you are human 2) same weapon type MH and OH 3) sword/fist if you are subject to a crappy computer or heavy lag.

Anonymous said...

Are you really sure about the "Sword spec will NOT proc off enchant/weapon/poison procs." part?

It seems to me that many people were getting the dual Lifestealing enchants to benefit the sword spec at most. Even though I don't have those enchants right now (dual agi), instant poisons seem to proc sword spec as well as kick sometimes.

Anyway I hope you'll continue to have a lot of fun w/ our next review & calculations ;)

Thanks for your blog,
Greg

Anonymous said...

I'm currently wielding an Ancient Qiraji Ripper with a Chromatically Tempered Sword (which incidentally makes me upset after reading the above).

I've tried some calculations on white dps only for dps increase and I'm wondering if these are correct since I'm not really an expert on numbercrunching.

Sword spec has a 5% chance of an extra MH attack, so it increases main hand dps. Taking the formula in your post you'd get something like this:

Normal mainhand attacks: 100% effectiveness so a 5% increase on main hand dps

Offhand attacks:
= [ 1 -(2.6/2.8)] * 5% increase of mainhand dps = 0.35% increase in main hand dps

Special attacks:
= [ 1 -(2.8/4)] * 5% increase of mainhand dps = 1.5% increase in main hand dps

Total = 6.85% increase in main hand dps

Say you take a fast offhand with a 1.8 speed as in your example, you'd then have a 8.3% total increase on main hand dps.


Considering this it wouldn't be that hard to calculate the difference between sword spec and fist spec.

Anonymous said...

using the sword spec with additional hit, would you say that the +hit gear you would have the better the spec would be compared to fist? With new PvP set one can get alot of +hit gear to push one self up tp +-18% hit. But doingthis you sacrifice +-50-70 AP? Is it worth it? For fist and sword or only sword? Is there a significant difference in DPS between 14% hit and 18% hit. Im struggling to set my self the correctly balance with sword spec.

BTW Your work is great and i really apreciate it!!

Luciano Tolomei said...

I'm going off topic.
currently i'm using a brutality MH and a Ravenholdt Slicer OH.

i'm specced 20/31 and i'm using
i was tring to model if it would be better something like:

hand of giustice / maelstorm card
and double life steal enchant.

or:

trinket 2%crit /trinket 2% hit
15 agility OH and a crusader MH

currently i'm at:
agy 360
AP 699
hit: 8% (without the weapon spec)
without any buff

VitaminC said...

Greg: I believe they nurf procs of procs so that sword spec won't proc off poison/enchant/weapon procs anymore.

3 posts above: Yes, a maladath OH will be better than a CTS OH in raid situations, as you lose less MH swings and gain more DPS on glancings (about 30-40% of your total white swings).

No, its still hard to compare as sword spec also introduce additional procs (crusader and poisons), which could make a big difference. Your poison DPS would also be increased by 5% from sword spec. On top of that is your additional weapon enchant procs.

Fist/dagger spec's dps increase is much more easier to calculate.

2 Posts above: Yes, it would be way better to have 18% hit compare to 14% hit even if you lose 70ap. You can gain AP quite easily in raids. And here is the thing, the higher your +hit is, the more sword spec proc you will have.

Luciano: I believe you will do more DPS with HoJ + 2% Hit trinket with MH Crusader and OH +15 agi, at least for raid PvE.

Anonymous said...

"Similar to OH proc modeling, assuming you aint doing ANYTHING STUPID like SnD/Evis/Rupture."
-what do you exactly meen :)?

VitaminC said...

Forgot why I wrote those, but basically I assumed only SS is used in the whole fights, which is why I use /4 at the denominator.

AFAIK, Evis/rupture will proc sword spec too but at a lower denominator (different than 4).

Also, SnD will change your effective weapon speed thus slightly change the effective sword spec proc from special attacks. SnD will not change your effective sword spec damage from OH though, as both MH/OH are increased at the same rate.

Anonymous said...

So if I have claw of the black drake and maladath, drake fang talisman + HoJ.

Currently 20/31 fist spec.

Would it be better for me to respec to sword spec for the OH sword triggering another MH claw strike? And then just save up DKP for CTS....

Anonymous said...

VitC 2 posts up:
lso, SnD will change your effective weapon speed thus slightly change the effective sword spec proc from special attacks. SnD will not change your effective sword spec damage from OH though, as both MH/OH are increased at the same rate.

True.
But, because the change is a rate, the actual change in each speed will be different. Suppose it is 25% faster, MH speed 10, OH speed 1 (This makes the difference easier for my math-challenged head to compute).
Pop Speedincreaseeffect. New MH speed 7.5, OH .75. Old MH/OH difference = 9, new =6.75. This difference in comparative speeds, because it fires off unpredictably in the fight, changes the way that the MH swing timer and OH swing timer sync up. When you begin fighting, the OH timer and MH timer start together. But from there, at MHspeed x OH speed (and possibly sooner) the two weapons are synced. At this point, a OH proc is exactly the same as a MH proc. But when you increase attack speeds, you change when exactly these sync up. In fact, because it is unclear when exactly increased attack speed factors in (suppose, above, we are 4 seconds through our 10 second swing timer. pop speed increase skill. do we have 5 seconds left, 6 seconds left, or something in the middle?), depending on when SND is popped, it changes sword spec dps. In any case, this is pretty much quibbling, but if one is to provide a totally accurate model, the way in which swing speed is changed mid swing-sequence is important. But then again, if one totally took this into account, I can't imagine it changing the dps at all really.

Zathras said...

In my addition to my above post I'd like to add that in order to calculate sword proc dps you'd have to take the sword procs off the sword procs and the procs of those as well.

So in my example it would be 6.85 * 1.0525 = 7.21% increase in main hand dps.

Thing is since this is extra procs, you can just calculate poison dps on the same modifier, so add 7.21% to your main hand (instant) poison damage.

Not entirely sure how you'd calculate crusader procs as they work on a ppm basis rather than a pps, though I'm sure I've seen somewhere that uptime is influenced by weapon speed.

In any case, if you use a modifier of 7.21% of extra attacks with a delay of 0, you could simply decrease the weapon speed by that amount in order to get the average weapon speed used for ppm. So where a 2.8 speed weapon gives you about 21.5 swings per minute, a 7.21% increase would make that 23 swings per minute, making the average weapon speed 2.6, which can then be used to calculate the new crusader procrate.

Rain (Frostmane EU) said...

One little questioin: I have seen you using the "attack-in-motion" term several times, but could you explain what exactly does it mean?

VitaminC said...

Shoulda be more specific. :(

What I meant by attack-in-motion is the time in between attack. i.e., you swing the sword in time 0:00, and the sword has speed of 0:05, than at 0:03 you are in "attack-in-motion". You will lose the 0:03 attack cooldown time if your OH procs a sword spec.

Rain (Frostmane EU) said...

Oh, now I see! Thank you!
And thanx for all of your articles - I like all that math, theory and statistics a lot!

I hope you don't mind if I translate some of your posts into Russian to put them on my guild forum, do you? (With link to your blog, of course!)

VitaminC said...

go ahead yo, aint all my work either. just compiling what i have learned in game and on US rogue forums.

Rav said...

Interesting, so sword spec isnt all its cracked up to be. That said, how is it better than claw spec? I thought CotBD rogues usually top dmg meters, losing out only if there is a Thunderfury in the raid?

Anonymous said...

You're right about the basic mechanic of sword spec, but your equations for cases 2 and 3 are quite wrong. They would only be true if there's a sword spec proc on the first offhand/instant attack after the main and offhand swings are synchronized, meaning right at the beginning of combat.

The better model is simpler and much more accurate. Basically, your offhand swings will basically occur at random distribution of times relative to main hand swings as the fight goes on. It's trivial to realize that offhand swings will occur halfway through the mainhand swing on average (assuming a flat distribution, which will be the case as fights become long, though events that resync main/off or main/instant timers will very very very slightly tilt the timer in the unfavorable direction). Therefore on average each proc from an offhand or instant swing is worth 1/2 of the intended value, and mainhand swings are still worth the full intended value.

This means that faster offhand weapons are better, that more instant attacks are better, and that (if you go look at the math for a pattern of attacks used in a typically long fight), sword spec does about 3% or what it should be doing, and is worth about 60% of fist or dagger spec's value.

Anonymous said...

So from doing out some math myself I came to realize you want either 1.) 2 swords of the same speed or 2.) an off hand with double the attack rate of the main hand. To maximize DPS. Otherwise over a long fight the OH procs will fall to towards a 50% average over time, ariving there at most 3 min later.

Anonymous said...

I've been concidering in using MH HWL's MH Fist, OH HWL's Quickblade
and going 5/5 swordspec, 5/5 fist spec.
After doing some testing I'll probably respec and go for dual swords, even if it's just for those oneshot lucky 8xproc's.
I'm curious tho if using equal weapons would fistspec be better.

Anonymous said...

I've been concidering in using MH HWL's MH Fist, OH HWL's Quickblade
and going 5/5 swordspec, 5/5 fist spec.
After doing some testing I'll probably respec and go for dual swords, even if it's just for those oneshot lucky 8xproc's.
I'm curious tho if using equal weapons would fistspec be better.

Anonymous said...

Hmm... either I don't understand your approach or it's flawed.

Main and Offhand swing timers are independent right? So when your offhand swings your mainhands swing-timer can be anywhere between 0 and mh_speed, correct? That means on an offhand proc you lose half the swing timer on average -> the offhand proc does 0.5 x MH damage.

The same should be true for specials.

Thus the final table should be independend from the weapons speed:

MH Sword Spec Proc = 100% of MH damage
OH sword spec proc = 50% of MH damage
Special sword spec proc = 50% of MH damage

Now its only a question what the impact of each source on the average proc-damage is and that depends of course on weather you use S&D and how fast your offhand is. All in all I'd guess sword spec is worth about 70% of the old sword spec where no swing timer got reset.

That would have been my approach: what did I do wrong?

Anonymous said...

I understand your results are likely from some parsing, however I thought blizzard was attempting to make the proc not reset the timer? Do they still continue to fail miserably? If so, do you have any links with the results of a long series of parsed data?