Attack speed vs Attack time

[tanis83]

Guys, I need some clarification on this. I just can't get the math right here.
If I have a 130% attack speed, should that decrease the attack time by 30%?

Theoretical example:
I equip an ordinary Tulwar sword with my warrior which results in:
Attack speed: 100%
Attack time: 640

Now how should Attack time change if I equip an armor for instance, with "Increased attack speed +30%"?

Based on my understanding it should change to 640*0.7=448, but this is definitely not the way it works.

[Britnoth]

No, +30% speed is not the same as -30% time between attacks.

Otherwise +100% speed would result in zero time between attacks.

0.640s at +30% speed is 640/1.3 = 0.492s

[tanis83]

Thanks for the info @Britnoth. It seems though the formula is still not precise enough.

Example: Chernobog's Claw is a Tulwar with a +30% increased attack speed. Equipping it shows:
Attack speed: 130%
Attack time: 544

What is going on here?

[jcvmarques]

If it's like D2, breakpoints are used to determine the frames necessary to do the action.

[tanis83]

@jcvmarques: I'm more interested in the math not the rendering part. Why is attack time 544 equipping a Tulwar with 30% increased attack speed for my warrior?

[Hassadar]

That could be it, in D2 u head breakpoints at certain FPS making it so that one would need to get over a certain amount of IAS (also works for cast speed) to get to the next faster attack animation. Everything added in between was just wasted.

Edit: eg.: u need more IAS to get over the next FPS breakpoint in order to see a lower attack time in the char menu

[tanis83]

I think I get the idea. And this correlates well with what I've seen when experimenting with items of IAS: in most cases attack time is a factor of 16, as if 16 was some kind of "stepping".

I can accept this (well, I have to), but I'm not too happy with the mystical calculations involved.

Oh well, it's still a great game if somewhat flawed here and there.

[Hassadar]

In D2 there were breakpoint tables that showed how much IAS/ICS one would need to reach the desired breakpoint. Of course the last breakpoint was almost impossible to reach and not practical (too much sacrifices to make to reach the breakpoint). Unfortunately in belzebub we don't know those breakpoints.

[tanis83]

Thanks @Hassadar for the clarification. I found the docs for D2 (there's even an attack speed calculator out there considering IAS). I now understand that IAS should not be considered as a percentage value, but more like a lookup value in a table where IAS ranges are mapped to frame (tick) counts. These frame cnts then represent the number of frames between attacks. 1 frame = 40 ms @25 FPS.

Not sure how similar the D1 engine works (let alone Belzebub), but could be very similar.

[Ritchie]

My thoughts on the IAS calc. 30% increase equates to 1+((1-1.3)/2) = .85. .85*640 = 544. Maybe a calc similar to this one is the one that is being used. Would have to do testing with other IAS values to confirm. It doesn't work with high values of IAS, but I'd assume there's a diminishing returns type deal. Where say a 100% increase wouldn't literally cut your attack time in half.

[Britnoth]

I just did a quick test, +100% speed does halve the attack time. Seems to be capped as 226% attack speed was no faster.

Also the intervals seem to be based on class, not the weapon. Very strange.

Makes attack speed rather weak unless you reach very high levels.

[Hassadar]

Base Attack speed from character class + bonus/malus for weapon type (bow/staff/axe etc) + base weapon speed from weapon class (ward bow/long war bow etc) + stats (overall dexterity) + IAS from gear = overall amount of attack speed which determines attack time. You need to reach certain attack speed breakpoints to get to the next better attack time. Idk how high these breakpoints are, but if you just reached the next better attack time, you would need A LOT more attack speed to get to the next better attack time.