Whirlwind Breakpoints

[Archon_Wing]

Ok, I have waded through all sorts of archives across many sites containing incredibly detailed analysis on the way our favorite Barbarian does the ballet. Yes, I do know only weapon IAS counts, breakpoints are at 10 and 35 for swords and 30/60 for two handed weapons. That's all fine and dandy, but there's just one problem:


What the heck does a breakpoint do? Should I bother with it?

Edit: err... -10/-35/ and -30/-60

[Elric of Grans]

Hail Archon Wing,

That seems odd: the base weapon speed (and class) does not count? Those figures do not sound right to me, though I must admit that I do not know the Whirlwind breakpoints myself.

What do they mean? It's the IAS required to lower the number of frames between attacks; the lower this number, the more attacks you make. It's not a bad idea to increase your speed so that you can make more regular attacks, though it is quite viable not to if that is your preference.

[Archon_Wing]

Oops, I meant weapon speed. And those speeds should be negative:)

[Guest]

Supposedly at each break point you get more attacks per whirle. So if its true it matters greatly.


Frankly I have always suspected this is false. Long long ago when LOD first came out, back on the old lounge I did tests on this and it seemd that I was getting a more linear result rathher than distict steps.

BUT - I dont know. It was too long ago for me to be sure.

I do know from more recent exp that my Hone Sudan with 3 shaels(-60) hits MUCH MUCH MUCH more than a plain Spire of Honor(+20) :)

[Bullit]

The breakpoints for 1 handed weapons are -10, and -35.

The breakpoints for 2 handed weapons are -10, -30, and -60.

The value for this is the WSM speed minus the total IAS on the _weapon_.

For example:

A colossus sword has WSM 5, so to hit the second 1H breakpoint, you need a total of 40% IAS (5-40 = -35) on the weapon.

edit - forgot to mention the purpose of the breakpoints :huh:

Hitting each one will result in more hits per whirlwind, which means more damage - always good ;) I believe the total hits takes the range of the weapon you're using into account too, so you might want to consider that.

[Elric of Grans]

Hail Bullit,

That's what looks wrong to me:

"A colossus sword has WSM 5, so to hit the second 1H breakpoint, you need a total of 40% IAS (5-40 = -35) on the weapon."

Why is it that with Whirlwind it directly modifies IAS with the weapon's modifier, when with the rest of the game they are completely separate figures? Can someone explain that to me?

[Dagni]

It's probably something stupid, like wereform attack speeds. I wouldn't be surprised if non-weapon IAS can make a very small difference, just like it can with wereform attacks. But I've neither tested nor read anything about this, so my guess is only idle speculation.

- Dagni

[Elric of Grans]

Hail Dagni,

I know that OIAS has no effect: that was intentional code. During the beta there was a bit of an outcry to the way Blizzard proposed to make Whirlwind work with weapon speeds, so they made it so OIAS had no effect at all; only WIAS. However, I would have thought it would not merely add WIAS and the WSM together: that seems rather stupid...well, actually, it makes more sense to me this way, but given it doesn't do that anywhere else, the lack of consistency seems stupid ;)

[Guest]

Well since its working with a completely differnt formula than other IAS calculations(#hits vs. attack speed), its seems reasonable that a seperate simpler code would be used.

[Hammerman]

This is my understanding of whirlwinds attack rate... but I'm not 100% sure of all of this.

Every 4th frame of the sequence, the game checks if it should make an attack. For the first time (frame 4 in the sequence), you will get an attack. The second time (frame 8), you also get an attack. After that, you get an attack based on the weapons ias/wsm:

delay = a1_frames*256 / ((weapon ias+wsm+100)*anim speed/100)

a1_frames = frame length of normal attack (attack1)
anim speed = 256 for all attacks
weapon ias = sum of fast/faster/fastest attack speed mods on the weapon
wsm = negative 'speed' from weapons.txt

This is then "breakpointed":

delay = real delay
0 = 4
12 = 6
15 = 8
18 = 10
20 = 12
23 = 14
26 = 16

If dual weilding, it changes wich weapon to use each successful attack. As before, the game checks every 4th frame if it should make an attack, and if the delay has passed you will get an attack.

bit ot: The formula I gave for 'delay' is also used to determine the very slow-very fast attack speed display on weapons (and it is also used in wereform attack speed calculations):

very slow attack speed = 28 or higher
very fast attack speed = 9 or lower

The rest of the variations (fast, slow, normal, etc) are different for char/weapon.

[Guest]

I really dont follow how that determins the number of "swings" per whirl.

It looks like you just have the normal attack speed calculation from pre LoD and the just say "this is then brakpointed".

Its quite likely I am all wrong, but you you explain that more clearly.

[Hammerman]

Sorry, I'll try to explain it better (and this is most certainly not pre-lod info).

As I said, every 4 frames the game checks if it should make an attack. The first 2 times (4th and 8th frame) you will get an attack. After that, there must pass at least some amount of frames, calculated as:

delay = a1_frames*256 / ((weapon ias+wsm+100)*anim speed/100)

There are breakpoints on this, sort of, as you can see from the table in my previous post. Let's say you get 20 as 'delay'. That means that next attack will be after 12 frames. If you're dual weilding, the game will switch between weapons every attack.

I hope this was clear enough. Oh and there is probably some more to this, this is at least the basics.

[Ruvanal]

Interesting. And it also took me a couple of minutes to realise how the "delay+real delay" fit into what you were talking about. I perfer this over some imperically derived values from a site where I do not know what testing methods that they used.

I plugged the formulas into a spread sheet and ran it through a series of Animation Frames and IAS factors.

For those that do not have the access to look up some of the Animation Frames, here are some for the barbarian.
a1 action mode
1hs=16 most 1 handed weapon use this
2hs=18 two-handed swords
2ht=19 two-handed spears
bow=15
hth=12 fists/no weapon in hand
stf=19 staves, polearms and the two-handed axes/hammers
xbw=20

Oddly I did not see "1ht" animation files for the barbarian as these would cover things like the knifes and javelin class weapons. Maybe I will check on that later.

edit: 1ht=16 knives and javelins class weapon in melee. Apparently shares the 1hs animations for all bodyparts except the arms and hands(weapon graphic).

In the formula the "weapon ias+wsm" factor can be treated as a single net factor of WIAS-WSM (faster weapons have a negative value so they will be quicker). This also assumes that I do not do any typos in transfering information from the spreadsheet. For the third a later attack you will get frame delays of the following based on the weapon used and the net WIAS-WSM factor.

1hs & 1ht
frames WIAS-WSM
12 -20
10 -19 to -11
8 -10 to +7
6 +8 to +33
4 +34 and up

2hs
12 -20 to -10
10 -11 to 0
8 +1 to +20
6 +21 to +50
4 +51 and up

2ht or stf
14 -20 to -18
12 -17 to -5
10 -5 to +5
8 +6 to +26
6 +27 to +58
4 +59 and up

hth
6 0
4 1+ (but fists are 0 and how do you get WIAS with no weapon?)

You cannot Whirlwind with bows and crossbows, so these are skipped.

bit ot: The formula I gave for 'delay' is also used to determine the very slow-very fast attack speed display on weapons (and it is also used in wereform attack speed calculations):

very slow attack speed = 28 or higher
very fast attack speed = 9 or lower

Hmm for the very fast and very slow attack speeds are you referring to states like when the character is chilled or speed up by some of the skills like Frenzy or Burst of Speed?

Also for the Druid wereform attacks, I would assume that it is using the human form a1_frames? If so I can run the tables out for these also since I have all the raw data at hand.

edit: forgot to exend the tables to cover weapons that have +10 and +20 WSM (slow).

[Hammerman]

hth
6 0
4 1+ (but fists are 0 and how do you get WIAS with no weapon?)

If you're not using a weapon, the "real delay" will be 10.

Hmm for the very fast and very slow attack speeds are you referring to states like when the character is chilled or speed up by some of the skills like Frenzy or Burst of Speed?

No, I was referring to the text on the weapons description, that says fast attack speed, normal attack speed, etc.

Also for the Druid wereform attacks, I would assume that it is using the human form a1_frames? If so I can run the tables out for these also since I have all the raw data at hand.

Oh there's a bit more to the druid wereform attack speed, that formula is just a factor in it. But yes, it uses the human form a1 frames. And after that, it divides the result ('delay') with previous animation length * 256 (wich should always be neutral)... lets call the result of that 'animation speed'. if using fists, then animation speed = 19.

Then:

item_ias = (item_ias * 120) / (item_ias+120) ; only if any ias
total ias = item_ias + 100 - speed + skill ias

Final animation modifier calculation is:

speed mod = (animation speed * ias) / 100

Tis is then used in normal animation display methods, wich is basically:

next_frame = (speed mod + (previous_frame*256)) / 256

[Brista]

For those that do not have the access to look up some of the Animation Frames, here are some for the barbarian.
a1 action mode
1hs=16  most 1 handed weapon use this
2hs=18  two-handed swords

Could one of you confirm which Animation Frame a 2H sword wielded in one hand would use, please?

Excellent work!

[Guest]

Your formula is giving delay.

But the issue that matters is total number of "swings" per whirl.


I think -

There is a hard cap on the number of swings for any given whirl.

This cap is infulenced by weapon speed, duel wieldding and possiblely other factors.


I know a cap exsists. Thats what all the game tests people do are looking for.

[Ruvanal]

For those that do not have the access to look up some of the Animation Frames, here are some for the barbarian.
a1 action mode
1hs=16   most 1 handed weapon use this
2hs=18   two-handed swords

Could one of you confirm which Animation Frame a 2H sword wielded in one hand would use, please?

Excellent work!

Two-handed swords that are dual wielded use the 1hs animations.

[AegisFang]

Hi guys, someone told me you guys were looking for some help with whirlwind speeds and numbers of hits.

From my playing experience and from wat i read here both matche perfectly, not saying everything is completely accurate to the letter, like for instance between the so called breakpoints you can get a little increase in whirlwind hits but its very small compared to what a breakpoint will give you still there is an improvement.

Barb faq dii.net
Upping the amount of hits whirlwind gives out is solely based on a system we call "break points", which work at specific IAS points. Max WW hits also seems to be capped according to your weapon range. Max hits per WW is Range + 3. So 6 maximum hits with a Col Blade, 8 with a War Pike.

Note : Applies to single opponents only. Mostly helpful if you are into dueling. Also, we do not know how this factors in with dual wielding whirlwind.
Second Note : I'm sure I will get bombarded with messages from mathematically inclined gamer, saying "you can't modify a weapon's base speed !" which is true. But for simplicity's sake, we'll use "speed" as in 10 base speed + 20 IAS via Shael = -10 speed. Butchering the math I know, but makes it a little easier to understand.

The so called break points here - weapon speeds that allow you to land more hits per one spin. The break points are:

-10, -30, -60 weapon speeds for 2h weapons
-10, -35 for 1h weapons and all swords.

A small example: a War Pike has weapon speed 20 (higher number means slower weapon). It means that a War Pike needs 30% IAS to hit the first break point, 50% IAS to hit the second BP and freaking 80% IAS to hit the last BP. All the IAS must be ON the weapon of course. Technically:

20 - 30 = -10 (1st BP)
20 - 50 = -30 (2nd BP)
20 - 80 = -60 (3rd BP)

Only a War Pike of Quickness with two Shael runes inside can reach the last break point.

Another example: a Colossus Blade has base speed 5 so it needs 15% IAS to hit the first sword break point at -10 and 40% IAS to reach the second and last BP at -35. Here's why:

5 - 15 = -10 (1st BP)
5 - 40 = -35 (2nd BP)

The last WW break point, i.e. -60 weapon speed for 2h weapons and -35 weapon speed for 1h weapons and swords can get you up to (weapon range + 3) hits per spin on a single, defenseless and immobile target.

That's why a War Pike with 80% IAS (-60 speed) can get up to 8 hits per whirl because it has range 5 and a Colossus Blade can get max 6 hits per spin because it has range 3 and 3 + 3 = 6

Also, the BP's for swords are universal whether they are used 1-Handed or 2-Handed. Please bear in mind that these tests were conducted on a solitary target, and due to the nature of WW, you can get many more hits if you whirl a longer distance or into mobs. There is a large difference between getting to a BP and not, especially the last ones listed, these will enable your weapon to hit at it's max rate (so the warpike above, if it didn't get to the -60 BP, would never ever get 8 hits per single target).


I hope this helps. :)

[whereagles]

Interesting.. the `free' 1st and 2nd hit does match what I experience in practice. WW feels bizarre and random in the hits, and this `free' hits thing coming the actual cracked code does make sense.

In practice the break point thing is all you need, but it is nice to know that even if you don't reach some of the break points, the weapon still hits a bit more. Some people claimed this was true, though not many believed them..

[Brista]

Hi guys, someone told me you guys were looking for some help with whirlwind speeds and numbers of hits.

Hi Aegisfang, have a closer look at the earlier replies from Hammerman and Ruvunal

They've gone into more depth than the breakpoint system