Welcome to Obsidian Forum Community

# Attack speed 2.0

176 replies to this topic

### #161 MasterCipher Posted 04 June 2016 - 07:52 AM

MasterCipher

(4) Theurgist

• Members
• 205 posts

Could use some help double checking my math. Goal is 0 weapon recovery without much going over much. Base recovery is the part I'm least certain on.

attack speed: the unlabored blade, durganized saber, outlander's frenzy

melee speed: vulnerable attack

two weapon: duel wield talent

dex: 21

weapon "haste" = (1.2*1.15*1.25) - 0.05 - 0.2 + 0.2 + 0.33 = 1.005

weapon recovery = baseRecovery - haste = 1 - 1.005 = -0.005 rounded up to 0 since there's no benefit for negative value.

### #162 MaxQuest Posted 04 June 2016 - 08:11 AM

MaxQuest

Arch-Mage

• Members
• 2360 posts

For main hand (with unlabored blade):

spd_coef = AttackSpeedMultiplier + DW + TWFighting + MeleeAttackSpeedMultiplier(VulnAttack) + ArmorSpeedFactor
spd_coef = (1.2 * 1.25 - 1) + 0.5 + (1.2 - 1) + (0.8 - 1) + (0.8 + 0.15 - 1) = 0.5 + 0.5 - 0.05 = 0.95 (or 1.95 old style)
rec_coef = max(0, 1 - 0.95) = 0.05
dex_coef = 1 + (21 - 10) / 33.3 = 1.33

(in frames)
base_action (dagger) (10 DEX) = 5 (delay) + 20 (attack duration) + 33 (recovery duration)
new_action (dagger) (10 DEX) = 5 (delay) + 20 (attack duration) + 1.65 (recovery duration)
new_action (dagger) (21 DEX) = 5 (delay) + 15 (attack duration) + 1.24 (recovery duration)

For off hand (with durganized saber):

spd_coef = AttackSpeedMultiplier + DW + TWFighting + MeleeAttackSpeedMultiplier(VulnAttack) + ArmorSpeedFactor
spd_coef = (1.15 * 1.25 - 1) + 0.5 + (1.2 - 1) + (0.8 - 1) - (0.8 + 0.15 - 1) = 0.43 + 0.5 - 0.05 = 0.88 (or 1.88 old style)
rec_coef = max(0, 1 - 0.88) = 0.12
dex_coef = 1 + (21 - 10) / 33.3 = 1.33

(in frames)
base_action (dagger) (10 DEX) = 5 (delay) + 30 (attack duration) + 50 (recovery duration)
new_action (dagger) (10 DEX) = 5 (delay) + 30 (attack duration) + 6 (recovery duration)
new_action (dagger) (21 DEX) = 5 (delay) + 22.5 (attack duration) + 4.5 (recovery duration)

Edited by MaxQuest, 04 June 2016 - 08:19 AM.

• AndreaColombo and MasterCipher like this

### #163 MasterCipher Posted 04 June 2016 - 09:53 AM

MasterCipher

(4) Theurgist

• Members
• 205 posts

Thank your for the quick reply! I had to stare at this for awhile, but the math makes sense to me now.

### #164 MaxQuest Posted 04 June 2016 - 10:00 AM

MaxQuest

Arch-Mage

• Members
• 2360 posts
It indeed can be confusing...

### #165 MasterCipher Posted 04 June 2016 - 01:46 PM

MasterCipher

(4) Theurgist

• Members
• 205 posts

spd_coef = AttackSpeedMultiplier + DW + TWFighting + MeleeAttackSpeedMultiplier(VulnAttack) + ArmorSpeedFactor

Is the DW value for a single weapon 0, and is it the same for 2 handed weapons?

Is the DW value for a and weapon+shield also 0?

### #166 MaxQuest Posted 05 June 2016 - 09:10 AM

MaxQuest

Arch-Mage

• Members
• 2360 posts
^ It's an additional coefficient that is added only if you dual-wield and do NOT hold a shield.
• MasterCipher likes this

### #167 Elric Galad Posted 05 June 2016 - 10:46 PM

(10) Necromancer

• Members
• 1569 posts
• Location:Paris
Stupid question MaxQuest : how does full attacks work for attack time and recovery time ?

I suppose it is faster than 2 attacks (no recovery between the 2 attacks) but it's hard to figure how much.
• peddroelm likes this

### #168 MaxQuest Posted 06 June 2016 - 01:23 AM

MaxQuest

Arch-Mage

• Members
• 2360 posts

^ It's not stupid at all.

From what I know at the moment:

- if you perform full-attack with dual-weapons, the recovery between two strikes is indeed skipped.

- you can perform full-attack only by using an ability. And in that case an array of speed effecting stuff does not apply. Specifically: DualWieldAttackSpeedMultiplier (those extra 0.5 if you DW), Single1HWeapRecovFactor (the bonus from durganized shield), MeleeAttackSpeedMultiplier (TWF talent) and RangedAttackSpeedMultiplier (i.e. malus from Penetrating Shot).

- Dex, AttackSpeedMultiplier, RateOfFireMultiplier and ArmorSpeedFactor still apply.

• peddroelm, Boeroer and MasterCipher like this

### #169 Elric Galad Posted 06 June 2016 - 01:53 AM

(10) Necromancer

• Members
• 1569 posts
• Location:Paris
So you get a free Vulnerable attack too ? Good to know.

Does it even depends on the weapon type ? (same for dagger and speed)
• MasterCipher likes this

### #170 Boeroer Posted 06 June 2016 - 01:57 AM

Boeroer

Arch-Mage

• Members
• 14943 posts
• Location:Bucharest, Romania

Good to know!

What about Riposte? It's an ability and does a Full Attack - but it's triggered passively while you do other stuff like running around. I think there's no animation whatsoever (it would totally mess with the walking/attacking animation that's played at that moment) - so I guess none of the speed/recovery rules apply? It just does two attack rolls and damage numbers pop up - that's it, right? Like with disengagement attacks?

### #171 MaxQuest Posted 06 June 2016 - 02:32 AM

MaxQuest

Arch-Mage

• Members
• 2360 posts

So you get a free Vulnerable attack too ? Good to know.

Seems so.

Does it even depends on the weapon type ? (same for dagger and speed)

Yeah, the recovery that follows after the full-attack should depend on the weapon type (i.e. it's base attack duration).

The thing is that the NotifyAttackComplete() method, the one responsible for the recovery duration calculation, computes the final recovery coefficient and multiplies it by totalTime which it gets as argument. That totalTime is equal to animation length of the attack phase (post dex).

In case of dual-wielding: theoretically totalTime could be the total animation duration for both swings; but it's more likely that instead of one call, that method is invoked twice. Once for first swing (with zero-recovery marker) and than for second swing. And the final recovery is only equal to the recovery of the weapon that hits last (minus those effects that do not apply). But I'd better test what is written in this last paragraph to be sure.

What about Riposte? It's an ability and does a Full Attack - but it's triggered passively while you do other stuff like running around.

Needs testing)

Edited by MaxQuest, 06 June 2016 - 02:34 AM.

• Boeroer and MasterCipher like this

### #172 MasterCipher Posted 06 June 2016 - 04:54 AM

MasterCipher

(4) Theurgist

• Members
• 205 posts

This makes for an interesting prospect: a weapon centric character could dump recovery reducing talents/items and load up on per encounter full attack abilities and have a lot of room for other talents and items.

### #173 Dr <3 Posted 06 June 2016 - 06:14 AM

Dr <3

(8) Warlock

• Members
• 1044 posts
Another quick question: after all, if you are at -80% recovery using dual wield, and in One hand you put a speed weapon (-20% recovery) you get:
A) -100% recovery with both weapons
Or B) -100% recovery with 1 hand and -80% with the other hand?

### #174 MaxQuest Posted 06 June 2016 - 06:30 AM

MaxQuest

Arch-Mage

• Members
• 2360 posts
^ B). Speed enchant does only affect the weapon (hand) it is on.
So you would have -80% on one hand, and -100-138% on the other (depending on what exact speed-increasing stuff did you use).

This makes for an interesting prospect: a weapon centric character could dump recovery reducing talents/items and load up on per encounter full attack abilities and have a lot of room for other talents and items.

If only there were so many full-attack abilities which you could continue spamming..
Hello Torment's Reach.
• Boeroer likes this

### #175 Dr <3 Posted 06 June 2016 - 07:17 AM

Dr <3

(8) Warlock

• Members
• 1044 posts
Tks maxquest!
Apart from monk also the rogue can pile up a good number af full attacks. But monks is clearly better. Anyway good old boeroer already did something similar in his wich doctor build
• Boeroer likes this

### #176 Elric Galad Posted 06 June 2016 - 08:00 AM

(10) Necromancer

• Members
• 1569 posts
• Location:Paris
Boeroer is not old.
He's just a guy with experience.
• Boeroer likes this

### #177 Boeroer Posted 06 June 2016 - 11:12 AM

Boeroer

Arch-Mage

• Members
• 14943 posts
• Location:Bucharest, Romania
I'd say both.

#### 0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users