d2jsp
Log InRegister
d2jsp Forums > Diablo II > Diablo 2: Resurrected > D2:R Discussion > Need Help Verifying Attack Speed Formulas D2r
Prev1456789Next
Add Reply New Topic New Poll
Member
Posts: 2,909
Joined: Oct 26 2003
Gold: 0.20
Nov 9 2021 03:24pm
Quote (gel87 @ Nov 9 2021 03:56pm)
I'll make an attempt later. I have some info to go from:

.


Sweet hope it works! Would be cool to see fendazon viable lol
Member
Posts: 7,655
Joined: Mar 6 2008
Gold: 15.00
Nov 9 2021 03:25pm
Quote (dumyguy @ 9 Nov 2021 22:16)
ah gee u solved everything thanks bro quality post


Np now donate thx
Member
Posts: 21,932
Joined: Sep 3 2006
Gold: 455.00
Nov 9 2021 04:59pm
Quote (gel87 @ Nov 9 2021 09:36pm)
Sorc
SOA11HS 20 256
SOA11HT 19 256
SOA12HS 24 256
SOA12HT 23 256
SOA1BOW 17 256
SOA1HTH 16 256
SOA1STF 18 256
SOA1XBW 20 256
SOA21HS 20 256

Sorc zealing, use same formula as pala.
She use same formula for normal attack as well, but:


What i found by this was that:
Normal attacks:
Hands = HTH-1 aka 15
Knife = HTH+1 aka 17 or maby trust-2 which also is 17.
Swinging weps = 1HS-2 aka 18

Zeal:
Frames serial = {[FramesPerDirection/2]/(1+(Eias/100))}
Frames normal attack = {FramesPerDirection / (1+(Eias/100))}
Zeal use listed framesPrDirection which is 20 for swords.

Werebear use regular formula, but u need to include:
Sorc's framesPrDirection.
Here animspeed is 256 anyway, so no change there like we needed to do with claw assa.


Zeal has a middle frame on sorc xD
So the 3 middle attacks has a rollback of 60%
Frames serial First= {[FramesPerDirection/2]/(1+(Eias/100))}
Frames serial Middle = {[FramesPerDirection*0,6]/(1+(Eias/100))}
Frames normal attack = {FramesPerDirection / (1+(Eias/100))}
Zeal use listed framesPrDirection which is 20 for swords.
Member
Posts: 21,932
Joined: Sep 3 2006
Gold: 455.00
Nov 10 2021 09:16am
There seems to be some difference in general in d2r.
So here is fcr formula:

EFCR = Floor(120*fcr/(120+fcr);1)
FramesPrCast = Ceiling(FramesPrDirection/(1+(Eias/100));1)-1

So assa has base of 17.
Assassin
Fcr - - EFCR - - Frames
0 - - 0 - - 16
8 - - 7 - - 15
16 - - 14 - - 14
27 - - 22 - - 13
42 - - 31 - - 12
65 - - 42 - - 11
102 - - 55 - - 10
168 - - 70 - - 9
345 - - 89 - - 8

So here everything hits old breakpoints all the way until u enter the 9 frame. This used to be 174.. Can any1 check if this is the case that its 168 now? :)

Amazon:
Frames pr direction: 20
Fcr - - EFCR - - Frames
0 - - 0 - - 19
7 - - 6 - - 18
14 - - 12 - - 17
22 - - 18 - - 16
32 - - 25 - - 15
48 - - 34 - - 14
68 - - 43 - - 13
99 - - 54 - - 12
151 - - 67 - - 11

No changes for amazon.

Barbarian
Frames pr direction: 14
Fcr - - EFCR - - Frames
0 - - 0 - - 13
9 - - 8 - - 12
20 - - 17 - - 11
37 - - 28 - - 10
60 - - 40 - - 9 (used to be 63)
105 - - 56 - - 8
200 - - 75 - - 7

Sorc
Frames pr direction: 14
Fcr - - EFCR - - Frames
0 - - 0 - - 13
9 - - 8 - - 12
20 - - 17 - - 11
37 - - 28 - - 10
60 - - 40 - - 9 (used to be 63)
105 - - 56 - - 8
200 - - 75 - - 7

Druid human form:
Frames pr direction: 15
But his animation speed is: 208
Ceiling(15*256/(208*(1+(Efcr/100)));1)-1

Fcr - - EFCR - - Frames
0 - - 0 - - 18
4 - - 3 - - 17
10 - - 9 - - 16
19 - - 16 - - 15
30- - 24 - - 14
44 - - 32 - - 13
68 - - 43 - - 12
99 - - 54 - - 11
157 - - 68 - - 10

Necro/paladin:
Frames pr direction: 16
Fcr - - EFCR - - Frames
0 - - 0 - - 15
8 - - 7 - - 14
18 - - 15 - - 13
30 - - 24 - - 12
48 - - 34 - - 11
75 - - 46 - - 10
120 - - 60 - - 9
223 - - 78 - - 8

This post was edited by gel87 on Nov 10 2021 09:52am
Member
Posts: 8,698
Joined: Apr 27 2009
Gold: 22,278.00
Nov 10 2021 09:29am
this test would be indeed a must to verify the non-round up in xx,00 frames for fcr in d2r
Member
Posts: 21,932
Joined: Sep 3 2006
Gold: 455.00
Nov 10 2021 10:31am
Oki, i can guess my way into action frames i think as well:
So assa has base action frame of 9 and a base frame of 17.

EFCR = Floor(120*fcr/(120+fcr);1)
FramesPrCast = Ceiling(FramesPrDirection/(1+(Eias/100));1)-1
Ceiling(9*256/(256*(1+(eias/100)));1)

We dont use -1 at the end of action frames.

Assassin
Fcr - - EFCR - - Frames - - action frame
0 - - 0 - - 16 - - 9
8 - - 7 - - 15 - - 9
15 - - 13 - - 15 - - 8
16 - - 14 - - 14 - - 8
27 - - 22 - - 13 - - 8
39 - - 29 - - 13 - - 7
42 - - 31 - - 12 - - 7
65 - - 42 - - 11 - - 7
86 - - 50 - - 11 - - 6
102 - - 55 - - 10 - - 6
168 - - 70 - - 9 - - 6
345 - - 89 - - 8 - - 6

Amazon
(there May be difference here depending if bow/jav/bear hands etc... But list stats 20 for all... But there are always devations which i check later)
Frames pr direction: 20
Fcr - - EFCR - - Frames - - action frame
0 - - 0 - - 19 - - 11
7 - - 6 - - 18 - - 11
11 - - 10 - - 18 - - 10
14 - - 12 - - 17 - - 10
22 - - 18 - - 16 - - 10
29 - - 23 - - 16 - - 9
32 - - 25 - - 15 - - 9
48 - - 34 - - 14 - - 9
56 - - 38 - - 14 - - 8
68 - - 43 - - 13 - - 8
99 - - 54 - - 12 - - 8
113 - - 58 - - 12 - - 7
151 - - 67 - - 11 - - 7

This post was edited by gel87 on Nov 10 2021 11:12am
Member
Posts: 21,932
Joined: Sep 3 2006
Gold: 455.00
Nov 10 2021 02:21pm
did a test with 60 vs 70 fcr barb to see if break is 60 or 63.
And it seems to be above 60 for me, so i guess its still 63...

Member
Posts: 8,698
Joined: Apr 27 2009
Gold: 22,278.00
Nov 10 2021 02:31pm
nice test gel!

yeah, seems at least this old breakpoint remains using round up eias for xx,00 frames, right?

if this happens with other classes we can assume only above old bps work with non-round up eias for xx,00 frames
( but we need more tests with other classes to corroborate this ^ )

i have some reports of ppl breaking the fcr limit, but no vid about fcr with non-round up eias for xx,00 frames
should be more easier to know with ias i assume
Member
Posts: 21,932
Joined: Sep 3 2006
Gold: 455.00
Nov 10 2021 02:51pm
Quote (ThoroBR @ Nov 10 2021 09:31pm)
nice test gel!

yeah, seems at least this old breakpoint remains using round up eias for xx,00 frames, right?

if this happens with other classes we can assume only above old bps work with non-round up eias for xx,00 frames
( but we need more tests with other classes to corroborate this ^ )

i have some reports of ppl breaking the fcr limit, but no vid about fcr with non-round up eias for xx,00 frames
should be more easier to know with ias i assume


Well even with zero kind of rounding its off.. Its wierd:
120*60/(120+60) = 40 efcr
14 / (1+(40/100))-1 = 9

120*61 / (120+61) = 40,44
14 / (1+(40,44/100))-1 = 8,97 (no rounding)

Now next eias is 41..
Which require 63 ias.
63 * 120 / (63+120) = 41,3 aka 41

Which would be below 9 frames and be rounded to 9.

So i guess next test will be 61fcr. To see how the game rounds fcr. If it rounds it at all on EFCR or not. This will give "data" to how it acts, and this data might be relevant for fhr/fbr/attack speed.

But wife is currently playing cs go on pc, so i guess i will quickscope some with kar98k on vanguard on xbox meanwhile xD
Member
Posts: 21,932
Joined: Sep 3 2006
Gold: 455.00
Nov 10 2021 02:55pm
Quote (molson_drinker @ Nov 9 2021 10:24pm)
Sweet hope it works! Would be cool to see fendazon viable lol


Hi, i had a go at it.
I currently Fail, but i will try to guess my way until it fits the old calc on the lower ias breaks.
There are so many scenario. Different rollbacks times. Does not fit the table.
Amazon skips 3 pictures when stabbing with spears, it skips 2 if no wep, there are 3 different animations, the animation is custom and i struggle to find a fit. I think the animation rate May be custom, but i dont find it listed. I try usual values like 208, but i failed. But i do this inbetween a daily rutine with kids, wife and stuff, so its rushed and i might find it later.

This post was edited by gel87 on Nov 10 2021 02:56pm
Go Back To D2:R Discussion Topic List
Prev1456789Next
Add Reply New Topic New Poll