Whirlwind Breakpoints

[whereagles]

Doh! I realized I got it all wrong. Upon rereading I think I got it now :)

So basically the `old theory' of (-10,-35) for 1h and (-10,-30,-60) is actually correct :D

[devzero]

1hs
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

This is very intresting as it contradicts the known formula for WW breakpoints, but still confirms them :)

1 hs breakpoints are 8 and 34, but are listed as 10 and 35. Academical diffrence since neither 8 nor 34 is atainable on a weapon, but confirms the formula.
2 hs breakpoints are 6, 27 and 51, but are listed as 10, 30 and 60. If the formula is correct this is a diffrence since you CAN reach 55%
2 ht breakpoints are 6, 27 and 59, but are listed as 10, 30 and 60. Again only an academical diffrence, but confirms the formula.

My bet is that most testing have been done on other stuff than swords wielded 2 handed, but testing at 50 and 60 with a 2 handed sword COULD give confirmation to a bp at 60.

There is also another question as for how often it hits. As i have understood WW is an area of effect skill. The aoe is decided by the range of the weapon beeing used. I think, from playing a ww baba quite a bit that WW dosnt hit just anything inside the aoe tho. Its more like it has 3 distinct directions it checks for things to hit. Ill try to illsutrate with this diagram. (this is not from any kind of testing tho, so i might be mistaken)

__X__
__B__
X___X

The question is WHERE does WW try to hit each time it checks? There are several options here and for all that i know it might check for any random monster within the aoe. I guess it only hits once every time tho so that you wont hit MORE monsters by ww'ing thru a herd compared to one monster, except for the fact that you have monsters inside the aoe more of the time.

Last question, using 2 weapons means supposedly that you hit 50% faster. Does that mean that it checks every 2 frames or instead of 4 what?

edit: changed diagram

[Guest]

Hammerins post has alot of detail - but I dont think it is relevent to this issue at all.

[Hammerman]

Hammerins post has alot of detail - but I dont think it is relevent to this issue at all.

By 'Hammerin' I assume you mean Hammerman.

How can my post not be relevant to whirlwind breakpoints? I gave information about how to calculate the delay between swings. Perhaps you should read my posts again and try to see the meaning behind them.

devzero:

Last question, using 2 weapons means supposedly that you hit 50% faster. Does that mean that it checks every 2 frames or instead of 4 what?

When dual weilding, the game will switch between default weapon each attack. so the first delay will be based on the weapon in the left hand (from the players perspective), second delay based on weapon in right hand, etc. And you will ofcourse attack with both weapons each attack.

[Ruvanal]

devzero:

This is very intresting as it contradicts the known formula for WW breakpoints, but still confirms them 

1 hs breakpoints are 8 and 34, but are listed as 10 and 35. Academical diffrence since neither 8 nor 34 is atainable on a weapon, but confirms the formula.
2 hs breakpoints are 6, 27 and 51, but are listed as 10, 30 and 60. If the formula is correct this is a diffrence since you CAN reach 55%
2 ht breakpoints are 6, 27 and 59, but are listed as 10, 30 and 60. Again only an academical diffrence, but confirms the formula.

My bet is that most testing have been done on other stuff than swords wielded 2 handed, but testing at 50 and 60 with a 2 handed sword COULD give confirmation to a bp at 60.

Thanks for articulating one of the points that I thinking about but did not put in the earlier post. Another point to compare with the testing that had been done before is that there is no indication of any testing at the slower speeds like using a Maul with no IAS on it. From my experience the sustained rate of attack after the first two hits is rather slow. But again this would not be something that most would bother to try to check on and difficult to test when using only a single target for the testing.

I have noted no directional bias in which monster is selected to carry out the attack against. I suspect that it will use a system similar to Thunderstorm for tracking which targets to attack. In that case it keeps a list of targets in range and adds/deletes form the list as they enter/leave. It will then just be cycling through the list for the next one to attack.

Last question, using 2 weapons means supposedly that you hit 50% faster. Does that mean that it checks every 2 frames or instead of 4 what?

Example: Using a Scimitar (WSM=-20) righthand and a Falcion (WSM=+20) left hand, with no IAS mods on them.
1st attack rh frame 4
2nd attack lh frame +4= 8
3rd attack rh frame +6= 14
4th attack lh frame +12= 26 horrible weapon :(
5th attack rh frame +6= 32
6th attack lh frame +12= 44
etc.
side note: Amost 2 seconds would have passed with only 6 attacks having been made, with the first 2 having happened in the first 1/3 of a second. Cases like this it would be better to just do very short dash WW and start again instead of a single continous WW attack.


Ghostiger:

Hammerins post has alot of detail - but I dont think it is relevent to this issue at all.

It is very relavent if you run the sequence out like I did above.

Example2: Hone Sudan (Yari, WSM=0) with 3 shaels(IAS +60)
1st attack is at frame 4
2nd attack frame +4= 8
3rd attack frame +4= 12 [/i]check the table for 2ht in my first post above[/i]
4th attack frame +4= 16
etc.
The sustained attack rate is 25/4= 6.25 attacks per second.

Example3: Spire of Honor(Lance, WSM=+20) WIAS-WSM= 0-20= -20
1st attack is at frame 4
2nd attack frame +4= 8
3rd attack frame +14= 22 [/i]check the table for 2ht in my first post above[/i]
4th attack frame +14= 36
etc.
The sustained attack rate after the first two attacks is 25/14= ~1.8 attacks per second or less than a third of what you should expect with the Shaeled Hone Sudan.

[AegisFang]

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

:huh: I can see that , i tried to make it simple for the members, going to deep in the equations sometimes will just loose the common guy who was just looking at the end results. But its always nice to see people going deeper into the secrets of diablo2 workings.

[Guest]

I tried to explain above.

If there is a cap(there seems to be) the is more important than the delay.

[Hammerman]

I tried to explain above.

If there is a cap(there seems to be) the is more important than the delay.

You mean a cap on hits per targets (the range +3 Aegisfang mentioned?)? There is no hard cap on it, but probably a logical one. With increasing range, you can start attacking sooner and end the attacking later (ie. you get in attacking range sooner and out of range later), effectively landing more hits. With lower range, you get into attacking range later, landing less hits.

Does this seem reasonable to you?

[Guest]

No one seems to be adressing if a cap exsists.

[Ruvanal]

No one seems to be adressing if a cap exsists.

There is no cap per se, but there is as Hammerman indicated a 'practical' cap based on various factors of how the mechanics work.

Normally WW will move at a set speed across the screen. This plus the weapon length and target size will determine how long the target is within the range to be hit. The shorter that the interval is between the successive attack checks, the more 'hits' that can be seen on a single target. It is this type of effect that seems to be what the other are talking about for the number of hits that they are getting. This would be better referred to as the maximum hits that one can get on single target than to be referring to it as a "cap". If you want to see more hits than the "caps" that they are talking about, try Whirlwinding down the length of a large Bone Wall (preferrably while WW-locked on a fast retreating target).

[Guest]

Ok I finnally get it. Thanks for bearing with me.

[Zath]

I am not sure if I understand what is being communicated in this thread, so I would like to try two sample calculations below and ask you to help me correct any mistakes.

==============================
Example #1: WW with a Giant Thresher that is socketed with one Shael.

a1_frames = 19
anim speed = 256
weapon ias = 20
wsm = 10

delay = a1_frames*256 / ((weapon ias+wsm+100)*anim speed/100) = 14.61

Looking at Hammerman's breakpoint table, a "delay" of 14.61 would correspond to a "real delay" of 8 frames. The sequence of events in one WW will therefore be...

Frame #0: Start WW
Frame #4: 1st free hit
Frame #8: 2nd free hit
Frame #12: 4 frames have passed since last hit. No hits here since "real delay" (8) has not passed.
Frame #16: 8 frames have passed since last hit. 3rd WW hit applied since "real delay" (8) has passed.
Frame #20: 4 frames have passed since last hit. No hits here since "real delay" (8) has not passed.
Frame #24: 8 frames have passed since last hit. 4th WW hit applied since "real delay" (8) has passed.
==============================
Example #2: Dual wield WW with Lightsabre in right hand and Doombringer in left
------------------------------
For Lightsabre we have...

a1_frames = 16
anim speed = 256
weapon ias = 20
wsm = 30

delay = a1_frames*256 / ((weapon ias+wsm+100)*anim speed/100) = 10.67

real delay = 6
------------------------------
For Doombringer we have...

a1_frames = 16
anim speed = 256
weapon ias = 0
wsm = 10

delay = a1_frames*256 / ((weapon ias+wsm+100)*anim speed/100) = 16.78

real delay = 10
------------------------------
The WW swquence is therefore...

Frame #0: Start WW
Frame #4: 1st free hit, with Lightsabre
Frame #8: 2nd free hit, with Doombringer
Frame #12: 4 frames have passed since last hit. No hits here since "real delay" for Lightsabre (6) has not passed.
Frame #16: 8 frames have passed since last hit. 3rd WW hit with Lightsabre is applied since "real delay" (6) has passed.
Frame #20: 4 frames have passed since last hit. No hits here since "real delay" for Doombringer (10) has not passed.
Frame #24: 8 frames have passed since last hit. No hits here since "real delay" for Doombringer (10) has not passed.
Frame #30: 12 frames have passed since last hit. 4th WW hit with Doombringer is applied since "real delay" (10) has passed.
==============================

Several questions I still have are:

1): What is one unit of WW? Or in other words, how can I tell when the WW sequences in the above samples will end and grant me two more free hits with more mana consumed?

2): Is it true that there is a cap on hits per WW at "weapon range +3"? Or is there a cap at all?

3): How do I read Hammerman's breakpoint table for "delay" to "real delay" conversion? If the "delay" is 5, should I go for 4 frame or 6 frame "real delay"?

4): Why are "real delay" values always even? Is this just how it is coded in game, or have the odd values been omitted?

5): Is there some point along the formula where I should round up or down?

[Ruvanal]

Zath:

delay = a1_frames*256 / ((weapon ias+wsm+100)*anim speed/100) = 14.61

Looking at Hammerman's breakpoint table, a "delay" of 14.61 would correspond to a "real delay" of 8 frames. The sequence of events in one WW will therefore be...

It should be...
delay = a1_frames*256 / ((weapon ias+wsm+100)*anim speed/100) = 14 (no fractions)

From Hammermans breakpoint table this 14 would yield a "real delay" of 6 frames.

The game in almost all cases will be using integer calculations (data storage and processing time are major reasons). In integer calculations there will not be any fractional value stored at any point of the calcualtions.

1): What is one unit of WW? Or in other words, how can I tell when the WW sequences in the above samples will end and grant me two more free hits with more mana consumed?

You have sort of answered your own question. Basically it is "one unit of WW" if you would not have needed to click again to have continued attacking. Try it some in a safe area without any mana leech to get the feel for it. Normally it not too long of a distance that you will travel unless you accidently click on a fleeing target (oh, what fun Whirlwinding after a fallen across the Stoney Fields).

2): Is it true that there is a cap on hits per WW at "weapon range +3"? Or is there a cap at all?

Answered for you by Tharn in this thread.
http://www.theamazonbasin.com/d2/forums/in...ST&f=54&t=25873

5): Is there some point along the formula where I should round up or down?

Yes, as I mentioned the game is using integer calculations. Here is the "delay" calculation that I used in a spreadsheet for example.

Code:

=INT(B$2*256/INT(($A3+100)*256/100))

where B$2 is a1_frames and $A3 is the WIAS-WSM factor.

3): How do I read Hammerman's breakpoint table for "delay" to "real delay" conversion? If the "delay" is 5, should I go for 4 frame or 6 frame "real delay"?

4): Why are "real delay" values always even? Is this just how it is coded in game, or have the odd values been omitted?

This appears to be one of the numerous internal tables that are in the DLLs. As such they can potentially be very arbitrary. The value in the "real delay" would be used if your "calculated delay" is equal to or greater than the "tables delay" value. That is why the 14 delay is translated to a "real delay" of 6 and not an 8.

Here is an additional spreadsheet calcualtion for those that may be intereseted in checking my earlier tables.

Code:

=IF(B3<12,4,IF(B3<15,6,IF(B3<18,8,IF(B3<20,10,IF(B3<23,12,IF(B3<26,14,16))))))

B3 is a lookup of the cell output of the previous formula. This calculation was translating the "delay" value to the "real delay" in frames.

[Dagni]

And you will ofcourse attack with both weapons each attack.

Really?

Because if that's so, then isn't Whirlwind the only attack where you can attack twice as often with two (identical) weapons as with just one?

I guess I'm just surprised because, if this is true, I'd expect it to be more highly publicized.

- Dagni

[Zath]

I made a spread sheet to confirm Ruvanal's breakpoint table and came up with this:

1hs: 1 hand weapons
1ht: Knife/javelin in melee
2hs: 2 hand swords
2ht: 2 hand spears
stf: Staves, polearms, 2 hand axe/hammer

Code:

                    WIAS-WSM Values
Frames  (1hs & 1ht)  (2hs)  (2ht & stf)
 14             -38          -30        -26
                  -31          -22        -18
 12             -30          -21*       -17
                  -20          -10         -5
 10             -19            -9*       -4*
                  -11            0           5
  8              -10            1           6
                   -7           20         26
  6                8            21         27
                   33           50         58
  4               34           51         59
                    +            +          +

* Discrepancy with original table

p.s. Is it just me or are the code tags in this forum completely dysfunctional?

[Zath]

And I would like to echo Dagni's question as well. Are we sure that dual wielding results in both weapons connecting on each hit?

[Hammerman]

Really?

Because if that's so, then isn't Whirlwind the only attack where you can attack twice as often with two (identical) weapons as with just one?

I guess I'm just surprised because, if this is true, I'd expect it to be more highly publicized.

- Dagni

Yes, it does attack with both weapon (except the first time). However, you won't be able to hit the same target with both weapons.

[Zath]

Looks like the definition of "s/s barbs" will soon be changing from sword/shield to sword/sword when this is publicized then. I can already imagine dual Shael'd-Azurewraths (or IAS jewels) dishing out an average of 9,375 magic & cold damage per second of WW after the first hit. Even the best Berserker would have a hard time just matching the magic portion of that attack, let alone making up for the cold part. Ouch! :blink:

[Tub]

Hi

you may have heard my name before, but probably you haven't. So, I'm Tub, from diablo2.de, responsible for testing the whole assassin attack speeds, mercenary attack speeds (the one's that diabloii.net has stolen) and some other data that can be found in Heri's Weapon Speed Calc. That being said, I hope you trust my values, although this is my first post here and I have this big red "newbie" sign attached to my shirt ;)


A while ago I made some tests about ww-speeds. I didn't use any disassembler, but plain tests.

a short summary of my methods:
- modded weapons with 100 to 101 dmg, no str/dex-bonus, different basespeeds
- lvl 99 barb with lots of AR, high str/dex, 100% manaleech
- modded Wendigos with exactly 10k hitpoints, no defense, lvl 1 (giving 95% cth), no regeneration and no movement

I counted how many times I had to whirl through a wendigo until it died, usually killing 3 wendigos per setup. Due to the 95% cth cap the results were a little off by 1 or 2 ww's sometimes, but altogether it was pretty solid.


My conclusion was:

caps for swords (1h and 2h, same caps!)
-10, 8, 34

caps for 2h-weapons:
-5, 27, 59

There were some more caps above 20, but as they don't exist in normal d2, I didn't take the time to narrow them down.

btw, hammerman's formula seems to be correct, because at -100 the game crashed with a div by 0 error ;)


those caps are almost identical to the values Ruvanal posted, with a few exceptions:

- there was no cap for swords at -20/-19, although there has to be at least one cap between -20 and -40
- there were no caps for 2ht/stf at -18/-17 and +5/+6

What happened here ? If there was a cap between -20 and -19 on swords, the wendigo should have died a couple of ww's earlier. But it didn't. You might argue that 42 instead of 43 ww's per wendigo is faster, but the next cap jumped to 33-34 ww's, then 25, and past the cap at 34 I needed 17 ww's for a kill. 43 to 42 is nothing.
Results for the 2ht/stf caps are similar. There either are no caps, or there are caps that virtually don't matter.


now, second issue:

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.

so you're saying, that both weapons' speeds matter ? My tests said something completely different.

It is widely believed that only the speed of your left hand weapon matters (you sure know all those grandfatherghtsabre-barbs, and those funny dudes holding the lightsabre in their right hand ;)). On the other hand, it's also widely believed amongst barbs that the caps are 10/35 and 30/60, and I still see some of them telling me that 9 fps is max frenzy speed.
I don't know if they have been hit on the head too often or if they're just really dizzy from all the whirling, but I learned to be suspicious about any information from that character class. That's why I initially started my ww speed tests ;)


so.. same methods as before, except I modded some new weapons with just 1 dmg. This way I could measure both weapon speeds seperate.

my tests didn't turn out too well, and I assume that there are some "undocumented features" here. (ever noticed that in some situations using normal attack, you can make a barb attack slower if you exchange one weapon with a faster one?)

but, although most of the data was confusing, i'm pretty sure both weapons always have the same speed (so there is no difference if your left or your right sword does the damage, providing left and right weaponspeeds don't change).

In most cases this speed really was your left hand weapon speed! Any explanation for this ?




great work so far hammerman (as usual), but there are still some questions left..


greetings
Tub

[Guest]

In 1.10 remember physical damage will be greatly improved relatibe to magic damage (in hell).

Currently I have a barb with 2 cruel phases with a shael. he does ok with whirlwind(better than with his frenzy) but nits not much better than my Hone Sudan barb.

I think the good ticket on WW will be a range 5, 2 hand weapon with -60 ias. If you get a weapon averaging 200 damage like this its going to be powerful