Difference between revisions of "Amazon attack speed"

From Basin Wiki
Jump to navigation Jump to search
(Cut and pasted Attack rate section from Strafe page, some reorganisation)
(→‎Strafe: Expanded explanation of effect of next delay and client/server discrepancies on attack rate)
Line 2,069: Line 2,069:
  
 
==Strafe==
 
==Strafe==
−
[[Strafe]] always fires a volley of at least two missile attacks, whose combined length is the sum of the First and Last Frame sub-columns in the tables below. Lengths added by further missile attacks are in the Follow-up Frame sub-column, from left to right.
+
[[Strafe]] is an [[attack]] sequence which always consists of at least two [[missile]]s, whose length is the sum of the First and Last Frame sub-columns in the tables below; lengths added by additional missiles (due to additional targets or more minimum arrows) are in the Follow-up Frame sub-column, from left to right (the Last Frame sub-column is normally the combined length of the follow-up interval between the last two missiles and the end of the sequence animation).
  
  
−
Due to the existence of the bugged breakpoints and [[Next Delay|next delay]], effective [[attack rate]] against multiple and particularly single targets is not straightforward.
+
For example, when two arrows are fired by a [[bow]] with {{Mb|=0 EIAS}} (6|3|10), sequence length is 16 (6+10) frames (the first arrow is fired in frame 6, the second 3 frames later and the remainder of the animation is 7 frames); when 10 arrows are fired, sequence length is 40 (6 + (3*8) + 10) frames. Average [[attack rate]] against dispersed targets is sequence length/missiles, so the two arrow sequence averages 8 (16/2) frames per arrow while the 10 arrow sequence averages 4 (40/10) frames per arrow.
  
  
−
Arrows and bolts travel at the same speed, so it takes the same time for either to pass through a target. Effective attack rate is maximised by ensuring it takes as long as possible for each missile to pass through a target, allowing the maximum time possible for [[Next Delay|next delay]] activated by a previous arrow to expire: firing sequences are a fixed length, so maximising collisions per sequence results in more collisions and thus more hits and damage over time. However, while both always take 3 frames to pass through a [[size]] 3 target independent of firing position, it takes 1-3 frames for them to pass through a size 2 target dependent on firing position.
+
Strafe has a 4 frame [[Next Delay|next delay]], which prevents its missiles colliding with a '''single''' target more than once every 5 frames (this does not affect multiple ''dispersed'' targets since next delay will expire even when alternating between just two targets, although it may affect multiple targets when consecutive missiles have to pass through one target to reach another). This does '''not''' mean that once the follow-up interval between missiles is less than 5 frames, only alternating arrows will collide with the target: although arrows and bolts normally only make a single collision check, and this check cannot occur while next delay is active, it normally takes time for them to pass through a target and next delay can expire while this is happening.
  
  
−
[[File:Strafe.png|thumb|center|600px|Arrows or bolts take 3 frames to pass through a size 2 target centered on dark green [[Radius|sub-tiles]], 2 frames to pass through one centered on green sub-tiles and 1 frame to pass through one centered on empty sub-tiles. Effective attack rate against a single target is always maximised at 3 frames and minimised at 1 frame.]]
+
The time it takes for an arrow or bolt (they are the same speed) to pass through a stationary target depends on its [[size]]: it always takes 3 frames to pass through a [[size#Size 3|size 3]] target and a single frame to pass through a [[size#Size 1|size 1]] target, while it takes 1-3 frames to pass through a [[size#Size 2|size 2]] target depending on its position.
 +
 
 +
 
 +
[[File:Strafe.png|thumb|center|600px|Arrows or bolts take 3 frames to pass through a stationary size 2 target centered on dark green [[Radius|sub-tiles]], 2 frames to pass through one centered on green sub-tiles and 1 frame to pass through one centered on empty sub-tiles. Effective attack rate against a single target is always maximised at 3 frames and minimised at 1 frame, so position your [[Amazon]] accordingly.]]
 +
 
 +
 
 +
Effective attack rate is maximised by ensuring it takes as long as possible for each missile to pass through the target, allowing the maximum time possible for next delay activated by a previous missile to expire: attack sequences are a fixed length, so maximising collisions per sequence results in more collisions and thus more hits and damage over time. Since the '''vast''' majority of targets are size 2-3, what follows assumes it takes 3 frames for an arrow to pass through a single stationary target: black cells in the Missile row indicate frames in which no missile is passing through a target, black cells in the Collision row indicate frames in which next delay is active.
 +
 
 +
 
 +
;4 frame follow-up
 +
{{Number table}} width=100%
 +
|-
 +
!Frame
 +
!width=2%|0
 +
!width=2%|1
 +
!width=2%|2
 +
!width=2%|3
 +
!width=2%|4
 +
!width=2%|5
 +
!width=2%|6
 +
!width=2%|7
 +
!width=2%|8
 +
!width=2%|9
 +
!width=2%|10
 +
!width=2%|11
 +
!width=2%|12
 +
!width=2%|13
 +
!width=2%|14
 +
!width=2%|15
 +
!width=2%|16
 +
!width=2%|17
 +
!width=2%|18
 +
!width=2%|19
 +
!width=2%|20
 +
!width=2%|21
 +
!width=2%|22
 +
!width=2%|23
 +
!width=2%|24
 +
!width=2%|25
 +
!width=2%|26
 +
!width=2%|27
 +
!width=2%|28
 +
!width=2%|29
 +
!width=2%|30
 +
!width=2%|31
 +
!width=2%|32
 +
!width=2%|33
 +
!width=2%|34
 +
!width=2%|35
 +
!width=2%|36
 +
!width=2%|37
 +
|-align=center
 +
!Missile
 +
|colspan=3 {{IP|=1}}||{{E}}
 +
|colspan=3 {{IP|=2}}||{{E}}
 +
|colspan=3 {{IP|=3}}||{{E}}
 +
|colspan=3 {{IF|=4}}||{{E}}
 +
|colspan=3 {{IP|=5}}||{{E}}
 +
|colspan=3 {{IP|=6}}||{{E}}
 +
|colspan=3 {{IP|=7}}||{{E}}
 +
|colspan=3 {{IF|=8}}||{{E}}
 +
|colspan=3 {{IP|=9}}||{{E}}
 +
|colspan=2 {{IP|=10}}
 +
|-align=center
 +
!Collision
 +
| {{IP|=1}} ||colspan=4 {{E}}
 +
| {{IP|=2}} ||colspan=4 {{E}}
 +
| {{IP|=3}} ||colspan=4 {{E}}
 +
|
 +
| {{IP|=4}} ||colspan=4 {{E}}
 +
| {{IP|=5}} ||colspan=4 {{E}}
 +
| {{IP|=6}} ||colspan=4 {{E}}
 +
|
 +
| {{IP|=7}} ||colspan=4 {{E}}
 +
| {{IP|=8}}
 +
|}
 +
 
 +
 
 +
Thus when 10 missiles are fired at a single stationary target, only 8 (missiles 1, 2, 3, 5, 6, 7, 9 and 10) can actually collide with it.
 +
 
 +
 
 +
;3 frame follow-up
 +
{{Number table}} width=100%
 +
|-
 +
!Frame
 +
!width=2%|0
 +
!width=2%|1
 +
!width=2%|2
 +
!width=2%|3
 +
!width=2%|4
 +
!width=2%|5
 +
!width=2%|6
 +
!width=2%|7
 +
!width=2%|8
 +
!width=2%|9
 +
!width=2%|10
 +
!width=2%|11
 +
!width=2%|12
 +
!width=2%|13
 +
!width=2%|14
 +
!width=2%|15
 +
!width=2%|16
 +
!width=2%|17
 +
!width=2%|18
 +
!width=2%|19
 +
!width=2%|20
 +
!width=2%|21
 +
!width=2%|22
 +
!width=2%|23
 +
!width=2%|24
 +
!width=2%|25
 +
!width=2%|26
 +
!width=2%|27
 +
!width=2%|28
 +
!width=2%|29
 +
|-align=center
 +
!Missile
 +
|colspan=3 {{IP|=1}}
 +
|colspan=3 {{IP|=2}}
 +
|colspan=3 {{IF|=3}}
 +
|colspan=3 {{IP|=4}}
 +
|colspan=3 {{IF|=5}}
 +
|colspan=3 {{IP|=6}}
 +
|colspan=3 {{IP|=7}}
 +
|colspan=3 {{IF|=8}}
 +
|colspan=3 {{IP|=9}}
 +
|colspan=3 {{IF|=10}}
 +
|-align=center
 +
!Collision
 +
| {{IP|=1}} ||colspan=4 {{E}}
 +
| {{IP|=2}} ||colspan=4 {{E}}
 +
| {{IP|=3}} ||colspan=4 {{E}}
 +
| {{IP|=4}} ||colspan=4 {{E}}
 +
| {{IP|=5}} ||colspan=4 {{E}}
 +
| {{IP|=6}} ||colspan=4 {{E}}
 +
|}
 +
 
 +
 
 +
Thus when 10 missiles are fired at a single stationary target, only 6 (missiles 1, 2, 4, 6, 7 and 9) can actually collide with it. Whether or not attack rate improves with fewer collisions in less time varies depending on the EIAS breakpoint and number of missiles fired.
 +
 
 +
 
 +
A further complication is introduced by a bug resulting in discrepancy in length between the sequence displayed by the client and that actually executed on the server. When the sequence is shorter on the client, when a single sequence ends the Amazon is unable to move or execute a different attack until the slower server sequence has ended. However, when consecutive sequences are fired, the server sequence is interrupted when the client sequence ends, and the next sequence begins on both the client and the server.
 +
 
 +
 
 +
For example, a bow with {{Mb|=50 EIAS}} fires at 4|2|7 on the client and 4|3|7 on the server, resulting in a 10 arrow sequence lasting 27 (4 + (2*8) + 7) frames on the client and 35 (4 + (3*8) + 7) frames on the server. So after a single sequence, the Amazon will appear to have stopped firing after 27 frames but she will be unable to do anything else until the sequence actually finishes 8 frames later. Consecutive sequences will result in the server sequence being interrupted after 27 frames:
 +
 
 +
 
 +
{{Number table}} width=100%
 +
!width=20%|Arrow
 +
!width=8%|1
 +
!width=8%|2
 +
!width=8%|3
 +
!width=8%|4
 +
!width=8%|5
 +
!width=8%|6
 +
!width=8%|7
 +
!width=8%|8
 +
!width=8%|9
 +
!width=8%|10
 +
|-align=center
 +
!Frame
 +
|4||7||10||13||16||19||22||25||{{IF|=28}}||{{IF|=31}}
 +
|}
 +
 
 +
 
 +
The server sequence is interrupted before the last two arrows can be fired, resulting in an average of 3.375 (27/8) frames per arrow when fired at multiple dispersed targets; without the interruption, the average would be 3.5 (35/10). When firing at a single stationary target, only 5 of those 8 arrows (1, 2, 4, 6 and 7) can collide with it due to next delay, resulting in an average of 5.4 (27/5) frames per arrow; without the interruption, the average would be ~5.83 (35/6).
 +
<br>
 
<br>
 
<br>
−
 
 
===BOW===
 
===BOW===
 
{{Number table}}
 
{{Number table}}

Revision as of 18:05, 25 July 2013

Diablo II Notation Basic Class Skill Item Ally Quest Area Monster Tale
Attribute Unit Animation Attack Damage Death
Slows Target Freeze target Freeze length Half Freeze Not Frozen Stun length Knockback Walk/run FR/W
Attack speed Weapon IAS Block rate FBR Cast rate FCR Hit recovery FHR
Amazon Assassin Barbarian Druid Necromancer Paladin Sorceress
  • Find Weapon Speed Modifier of your weapon, then the table for your attack and weapon.


Increased Attack Speed only

  • Consult relevant WSM sub-column in IAS column to find target break-point.


EIAS

  • Consult EIAS column to find target break-point, then apply relevant modifiers:
Required EIAS = Target EIAS - Skill EIAS - chill - Slows Target + WSM
Skill EIAS = (Fanaticism + Holy Freeze + Decrepify)


Attack

1HT

These tables also apply to Power Strike, Charged Strike and Lightning Strike.


Frames EIAS IAS
20 10 0 -10
16 -23
15 -18 3
14 -13 8
13 -7 15 4
12 0 24 11
11 9 39 23 10
10 19 58 39 23 10
9 31 89 63 42 26
8 45 142 102 72 50
7 63 270 187 133 95


2HT

These tables also apply to Power Strike, Charged Strike and Lightning Strike.


Frames EIAS IAS
20 10 0 -10 -20
20 -23
19 -19 2
18 -15 6
17 -10 11
16 -5 18 6
15 0 24 11
14 8 37 22 9
13 15 50 32 18 6
12 24 70 48 30 16 5
11 34 99 70 48 30 16
10 46 147 105 75 52 34
9 61 250 174 125 89 63


BOW

These breakpoints also apply to all Bow and Crossbow Skills except Strafe.


Frames EIAS IAS
Total Action 10 5 0 -10
15 7 -12
14 -6 5
13 6 0 11 6
12 8 22 15 9
11 17 35 27 20 8
5 21 42 34 26 13
10 28 56 46 37 22
9 41 89 75 63 42
4 50 120 102 86 60
8 56 147 125 105 75
7 75 292 240 200 142


When Immolation Arrow is fired continuously, the interval between attacks is its 25 frame delay plus action flag (the frame of the animation in which the action is flagged to occur). [1]


Frames EIAS IAS
10 5 0 -10
32 -12
31 0 11 6
30 21 42 34 26 13
29 50 120 102 86 60


XBOW

These breakpoints also apply to all Bow and Crossbow Skills except Strafe.


Frames EIAS IAS
Total Action 10 0 -10 -40 -60
22 11 -12
10 -9 2
21 -8 3
20 -4 7
19 9 0 11
18 6 19 7
17 12 27 14 3
8 13 29 15 4
16 18 37 22 9
15 25 50 32 18
7 29 58 39 23
14 34 70 48 30
13 43 95 68 46 4
6 50 120 86 60 11
12 54 138 99 70 16
11 67 215 152 109 35 8


When Immolation Arrow is fired continuously, the interval between attacks is its 25 frame delay plus action flag (the frame of the animation in which the action is flagged to occur). [2]


Frames EIAS IAS
10 0 -10 -40 -60
36 -12
35 -9 2
34 0 11
33 13 29 15 4
32 29 58 39 23
31 50 120 86 60 11


Fend

  • Against a single target, Fend executes one attack the same length as an Attack.
  • Against two targets, two attacks whose combined length is the sum of the First and Last Frame sub-columns are executed.
  • Against 3-12 targets, the lengths of additional attacks are in the Follow-up Frame sub-column, from left to right.


1HT

Frames/Attack EIAS IAS
First Follow-up Last 20 10 0 -10
9 8 8 8 8 8 8 8 8 8 8 15 -21
14 -19 2
7 7 7 7 7 7 7 7 7 7 13 -14 7
8 -12 9
12 -7 15 4
7 6 6 6 6 6 6 6 6 6 6 11 0 24 11
10 10 40 24 11
5 6 6 6 6 6 6 6 6 6 15 50 32 18 6
6 6 5 6 5 6 5 6 5 6 5 17 54 35 20 8
5 5 5 5 5 5 5 5 5 5 9 21 63 42 26 13
4 5 5 5 5 5 5 5 5 5 8 34 99 70 48 30
5 5 4 5 4 5 4 5 4 5 4 41 125 89 63 42
4 4 4 4 4 4 4 4 4 4 7 50 168 120 86 60
6 61 250 174 125 89
3 4 4 4 4 4 4 4 4 4 7 67 317 215 152 109
6 72 395 259 180 129
4 4 3 4 3 4 3 4 3 4 3 75 456 292 200 142


2HT

Frames/Attack EIAS IAS
First Follow-up Last 20 10 0 -10 -20
12 9 9 9 9 9 9 9 9 9 9 17 -21
11 16 -17 4
8 8 8 8 8 8 8 8 8 8 15 -12 9
10 -9 13 2
14 -6 16 5
9 7 7 7 7 7 7 7 7 7 7 13 0 24 11
12 8 37 22 9
8 13 46 29 15 4
6 6 6 6 6 6 6 6 6 6 11 17 54 35 20 8
10 28 80 56 37 22 9
7 29 83 58 39 23 10
5 5 5 5 5 5 5 5 5 5 9 41 125 89 63 42 26
6 50 168 120 86 60 40
8 56 208 147 105 75 52
4 4 4 4 4 4 4 4 4 4 7 75 456 292 200 142 102


Impale

1HT

Frames EIAS IAS
20 10 0 -10
42 -20
41 -18 3
40 -17 4
39 -16 5
38 -14 7
37 -12 9
36 -11 10
35 -9 13 2
34 -7 15 4
33 -6 16 5
32 -4 19 7
31 -2 22 9
30 1 26 13 2
29 3 29 15 4
28 5 32 18 6
27 9 39 23 10
26 11 42 26 13 2
25 15 50 32 18 6
24 18 56 37 22 9
23 22 65 44 27 14
22 26 75 52 34 19
21 30 86 60 40 24
20 36 105 75 34
19 41 125 89 63 42
18 47 152 109 78 54
17 54 194 138 99 70
16 62 259 180 129 92
15 71 377 250 174 125
14 80 600 360 240 168
13 92 1,680 680 395 259
12 105 ∞ 2,760 840 456


2HT

Frames EIAS IAS
20 10 0 -10 -20
48 -20
47 -18 3
46 -17 4
45 -16 5
44 -15 6
43 -14 7
42 -12 9
41 -11 10
40 -9 13 2
39 -8 14 3
38 -6 16 5
37 -4 19 7
36 -3 20 8
35 -1 23 10
34 1 26 13 2
33 4 30 16 5
32 5 32 18 6
31 8 37 22 9
30 11 42 26 13 2
29 13 46 29 15 4
28 16 52 34 19 7
27 20 60 40 11
26 23 68 46 29 15 4
25 27 78 54 35 20 8
24 30 86 60 40 24 11
23 35 102 72 50 32 18
22 40 120 86 60 40 24
21 45 142 102 72 50 32
20 51 174 125 89 63 42
19 57 215 152 109 78 54
18 64 280 194 138 99 70
17 72 395 259 180 129 92
16 80 600 360 240 168 120
15 91 1,480 638 377 250 174
14 102 ∞ 1,680 680 395 259


Jab

The Average Frames sub-column is the average length of each of Jab's three attacks: each attack actually differs in length.

1HT

Frames EIAS IAS
Total Avg 20 10 0 -10
36 12 0/3 -20
35 11 2/3 -18 3
34 11 1/3 -16 5
33 11 0/3 -15 6
32 10 2/3 -13 8
31 10 1/3 -11 10
30 10 0/3 -9 13 2
29 9 2/3 -7 15 4
28 9 1/3 -5 18 6
27 9 0/3 -3 20 8
26 8 2/3 0 24 11
25 8 1/3 3 29 15 4
24 8 0/3 5 32 18 6
23 7 2/3 9 39 23 10
22 7 1/3 13 46 29 15 4
21 7 0/3 16 52 34 19 7
20 6 2/3 21 63 42 26 13
19 6 1/3 25 72 50 32 18
18 6 0/3 30 86 60 40 24
17 5 2/3 37 109 78 54 35
16 5 1/3 43 133 95 68 46
15 5 0/3 51 174 125 89 63
14 4 2/3 59 232 163 117 83
13 4 1/3 69 345 232 163 117
12 4 0/3 80 600 360 240 168
11 3 2/3 94 2,280 780 434 280


2HT

Frames EIAS IAS
Total Avg 20 10 0 -10 -20
42 14 0/3 -20
41 13 2/3 -18 3
40 13 1/3 -17 4
39 13 0/3 -16 5
38 12 2/3 -14 7
37 12 1/3 -12 9
36 12 0/3 -11 10
35 11 2/3 -9 13 2
34 11 1/3 -7 15 4
33 11 0/3 -6 16 5
32 10 2/3 -4 19 7
31 10 1/3 -2 22 9
30 10 0/3 1 26 13 2
29 9 2/3 3 29 15 4
28 9 1/3 5 32 18 6
27 9 0/3 9 39 23 10
26 8 2/3 11 42 26 13 2
25 8 1/3 15 50 32 18 6
24 8 0/3 18 56 37 22 9
23 7 2/3 22 65 44 27 14 3
22 7 1/3 26 75 52 34 19 7
21 7 0/3 30 86 60 40 24 11
20 6 2/3 36 105 75 52 34 19
19 6 1/3 41 125 89 63 42 26
18 6 0/3 47 152 109 78 54 35
17 5 2/3 54 194 138 99 70 48
16 5 1/3 62 259 180 129 92 65
15 5 0/3 71 377 250 174 125 89
14 4 2/3 80 600 360 240 168 120
13 4 1/3 92 1,680 680 395 259 180
12 4 0/3 105 ∞ 2,760 840 456 292


Strafe

Strafe is an attack sequence which always consists of at least two missiles, whose length is the sum of the First and Last Frame sub-columns in the tables below; lengths added by additional missiles (due to additional targets or more minimum arrows) are in the Follow-up Frame sub-column, from left to right (the Last Frame sub-column is normally the combined length of the follow-up interval between the last two missiles and the end of the sequence animation).


For example, when two arrows are fired by a bow with 0 EIAS (6|3|10), sequence length is 16 (6+10) frames (the first arrow is fired in frame 6, the second 3 frames later and the remainder of the animation is 7 frames); when 10 arrows are fired, sequence length is 40 (6 + (3*8) + 10) frames. Average attack rate against dispersed targets is sequence length/missiles, so the two arrow sequence averages 8 (16/2) frames per arrow while the 10 arrow sequence averages 4 (40/10) frames per arrow.


Strafe has a 4 frame next delay, which prevents its missiles colliding with a single target more than once every 5 frames (this does not affect multiple dispersed targets since next delay will expire even when alternating between just two targets, although it may affect multiple targets when consecutive missiles have to pass through one target to reach another). This does not mean that once the follow-up interval between missiles is less than 5 frames, only alternating arrows will collide with the target: although arrows and bolts normally only make a single collision check, and this check cannot occur while next delay is active, it normally takes time for them to pass through a target and next delay can expire while this is happening.


The time it takes for an arrow or bolt (they are the same speed) to pass through a stationary target depends on its size: it always takes 3 frames to pass through a size 3 target and a single frame to pass through a size 1 target, while it takes 1-3 frames to pass through a size 2 target depending on its position.


Arrows or bolts take 3 frames to pass through a stationary size 2 target centered on dark green sub-tiles, 2 frames to pass through one centered on green sub-tiles and 1 frame to pass through one centered on empty sub-tiles. Effective attack rate against a single target is always maximised at 3 frames and minimised at 1 frame, so position your Amazon accordingly.


Effective attack rate is maximised by ensuring it takes as long as possible for each missile to pass through the target, allowing the maximum time possible for next delay activated by a previous missile to expire: attack sequences are a fixed length, so maximising collisions per sequence results in more collisions and thus more hits and damage over time. Since the vast majority of targets are size 2-3, what follows assumes it takes 3 frames for an arrow to pass through a single stationary target: black cells in the Missile row indicate frames in which no missile is passing through a target, black cells in the Collision row indicate frames in which next delay is active.


4 frame follow-up
Frame 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
Missile 1 2 3 4 5 6 7 8 9 10
Collision 1 2 3 4 5 6 7 8


Thus when 10 missiles are fired at a single stationary target, only 8 (missiles 1, 2, 3, 5, 6, 7, 9 and 10) can actually collide with it.


3 frame follow-up
Frame 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Missile 1 2 3 4 5 6 7 8 9 10
Collision 1 2 3 4 5 6


Thus when 10 missiles are fired at a single stationary target, only 6 (missiles 1, 2, 4, 6, 7 and 9) can actually collide with it. Whether or not attack rate improves with fewer collisions in less time varies depending on the EIAS breakpoint and number of missiles fired.


A further complication is introduced by a bug resulting in discrepancy in length between the sequence displayed by the client and that actually executed on the server. When the sequence is shorter on the client, when a single sequence ends the Amazon is unable to move or execute a different attack until the slower server sequence has ended. However, when consecutive sequences are fired, the server sequence is interrupted when the client sequence ends, and the next sequence begins on both the client and the server.


For example, a bow with 50 EIAS fires at 4|2|7 on the client and 4|3|7 on the server, resulting in a 10 arrow sequence lasting 27 (4 + (2*8) + 7) frames on the client and 35 (4 + (3*8) + 7) frames on the server. So after a single sequence, the Amazon will appear to have stopped firing after 27 frames but she will be unable to do anything else until the sequence actually finishes 8 frames later. Consecutive sequences will result in the server sequence being interrupted after 27 frames:


Arrow 1 2 3 4 5 6 7 8 9 10
Frame 4 7 10 13 16 19 22 25 28 31


The server sequence is interrupted before the last two arrows can be fired, resulting in an average of 3.375 (27/8) frames per arrow when fired at multiple dispersed targets; without the interruption, the average would be 3.5 (35/10). When firing at a single stationary target, only 5 of those 8 arrows (1, 2, 4, 6 and 7) can collide with it due to next delay, resulting in an average of 5.4 (27/5) frames per arrow; without the interruption, the average would be ~5.83 (35/6).

BOW

Frames/Attack EIAS IAS
First Follow-up Last 10 5 0 -10
7 4 4 4 4 4 4 4 4 12 -14
11 -8 3
6 3 3 3 3 3 3 3 3 10 0 11 6
9 11 26 19 13 2
5 21 42 34 26 13
8 23 46 37 29 15
7 38 80 68 56 37
4 2 2 2 2 2 2 2 2 50 120 102 86 60
6 58 157 133 113 80


  • 50 and 58 EIAS breakpoints only apply to the displayed animation for follow-up attacks: they actually continue to be 3 frames apart, and this interval is never reduced to 2 frames. When Strafe is fired continuously, the next volley begins when the animation of the previous one finishes, which reduces the number of actual attacks per volley but may sometimes increase the number of actual attacks over time, resulting in a faster attack rate. [3]


Single target

The table below lists the average frames per arrow when firing continuously at a single target.


EIAS Frames/Attack Arrows
Server Client 2 3 4 5 6 7 8 9 10
-14 7|4|12 7|4|12 9.5 (19/02) 7.6 (23/03) 9.0 (27/03) 7.7 (31/04) 7.0 (35/05) 6.5 (39/06) 7.1 (43/06) 6.7 (470/7) 6.3 (51/08)
-8 7|4|11 7|4|11 9.0 (18/02) 7.3 (22/03) 8.6 (26/03) 7.5 (30/04) 6.8 (34/05) 6.3 (38/06) 7.0 (42/06) 6.5 (46/07) 6.2 (50/08)
0 6|3|10 6|3|10 8.0 (16/02) 9.5 (19/02) 7.3 (22/03) 8.3 (25/03) 7.0 (28/04) 6.2 (31/05) 6.8 (34/05) 6.1 (37/06) 6.6 (40/06)
11 6|3|09 6|3|09 7.5 (15/02) 9.0 (18/02) 7.0 (21/03) 8.0 (24/03) 6.7 (27/04) 6.0 (30/05) 6.6 (33/05) 6.0 (36/06) 6.5 (39/06)
21 5|3|09 5|3|09 7.0 (14/02) 8.5 (17/02) 6.6 (20/03) 7.6 (23/03) 6.5 (26/04) 5.8 (29/05) 6.4 (32/05) 5.8 (35/06) 6.3 (38/06)
23 5|3|08 5|3|08 6.5 (13/02) 8.0 (16/02) 6.3 (19/03) 7.3 (22/03) 6.2 (25/04) 5.6 (28/05) 6.2 (31/05) 5.6 (34/06) 6.1 (37/06)
38 5|3|07 5|3|07 6.0 (12/02) 7.5 (15/02) 6.0 (18/03) 7.0 (21/03) 6.0 (24/04) 5.4 (27/05) 6.0 (30/05) 5.5 (33/06) 6.0 (36/06)
50 4|3|07 4|2|07 5.5 (11/02) 6.5 (13/02) 5.0 (15/03) 5.6 (17/03) 5.0 (76/15) 5.2 (21/04) 5.1 (46/09) 5.0 (25/05) 5.4 (27/05)
58 4|3|06 4|2|06 5.0 (10/02) 6.0 (12/02) 5.2 (42/08) 5.3 (16/03) 6.0 (18/03) 5.2 (21/04) 5.0 (66/13) 5.3 (48/09) 5.2 (26/05)

[4]


  • Sequence length and thus average frames per arrow was found experimentally to be the same at the 50 and 58 EIAS breakpoint for 7 arrows (level 21-23) [5]


Multiple targets

The table below lists the average frames per arrow when firing continuously at multiple targets which are spread out so that no arrow has to pass through another target (thus even with two targets there's an interval during which the 4 frame next delay expires).


EIAS Frames/Attack Arrows
Server Client 2 3 4 5 6 7 8 9 10
-14 7|4|12 7|4|12 9.5 (19/02) 7.6 (23/03) 6.7 (27/04) 6.2 (31/05) 5.8 (35/06) 5.5 (39/07) 5.3 (43/08) 5.2 (47/09) 5.1 (51/10)
-8 7|4|11 7|4|11 9.0 (18/02) 7.3 (22/03) 6.5 (26/04) 6.0 (30/05) 5.6 (34/06) 5.4 (38/07) 5.2 (42/08) 5.1 (46/09) 5.0 (50/10)
0 6|3|10 6|3|10 8.0 (16/02) 6.3 (19/03) 5.5 (22/04) 5.0 (25/05) 4.6 (28/06) 4.4 (31/07) 4.2 (34/08) 4.1 (37/09) 4.0 (40/10)
11 6|3|09 6|3|09 7.5 (15/02) 6.0 (18/03) 5.2 (21/04) 4.8 (24/05) 4.5 (27/06) 4.2 (30/07) 4.1 (33/08) 4.0 (36/09) 3.9 (39/10)
21 5|3|09 5|3|09 7.0 (14/02) 5.6 (17/03) 5.0 (20/04) 4.6 (23/05) 4.3 (26/06) 4.1 (29/07) 4.0 (32/08) 3.8 (35/09) 3.8 (38/10)
23 5|3|08 5|3|08 6.5 (13/02) 5.3 (16/03) 4.7 (19/04) 4.4 (22/05) 4.1 (25/06) 4.0 (28/07) 3.8 (31/08) 3.7 (34/09) 3.7 (37/10)
38 5|3|07 5|3|07 6.0 (12/02) 5.0 (15/03) 4.5 (18/04) 4.2 (21/05) 4.0 (24/06) 3.8 (27/07) 3.7 (30/08) 3.6 (33/09) 3.6 (36/10)
50 4|3|07 4|2|07 5.5 (11/02) 4.3 (13/03) 3.7 (15/04) 3.4 (17/05) 3.1 (19/06) 3.5 (21/06) 3.2 (23/07) 3.1 (25/08) 3.3 (27/08)
58 4|3|06 4|2|06 5.0 (10/02) 4.0 (12/03) 3.5 (14/04) 3.2 (16/05) 3.6 (18/05) 3.5 (21/06) 3.1 (22/07) 3.4 (24/07) 3.2 (26/08)

[6]


  • Sequence length and thus average frames per arrow was found experimentally to be the same at the 50 and 58 EIAS breakpoint for 7 arrows (level 21-23) [7]


XBOW

Frames/Attack EIAS IAS
First Follow-up Last 10 0 -10 -40 -60
11 6 6 6 6 6 6 6 6 17 -10
10 -9 2
16 -5 6
9 5 5 5 5 5 5 5 5 15 0 11
14 8 22 9
4 5 5 5 5 5 5 5 12 27 14 3
8 5 5 5 5 5 5 5 5 13 29 15 4
13 15 32 18 6
5 4 5 4 5 4 5 4 21 42 26 13
12 24 48 30 16
4 4 4 4 4 4 4 4 11 25 50 32 18
7 12 29 58 39 23
11 34 70 48 30
3 4 4 4 4 4 4 4 43 95 68 46 4
10 46 105 75 52 7
6 4 3 4 3 4 3 4 3 50 120 86 60 11
9 61 174 125 89 26 2
3 3 3 3 3 3 3 3 8 67 215 152 109 35 8


  • >49 EIAS breakpoints only apply to the displayed animation for follow-up attacks: they actually continue to be 4 frames apart, and this interval is never reduced to 3 frames. When Strafe is fired continuously, the next volley begins when the animation of the previous one finishes, which reduces the number of actual attacks per volley but may sometimes increase the number of actual attacks over time, resulting in a faster attack rate. There are other breakpoints which only apply to the displayed animation rather than actual attacks as well. [8]


Single target

The table below lists the average frames per bolt when firing continuously at a single target.


EIAS Frames/Attack Arrows
Server Client 2 3 4 5 6 7 8 9 10
-10 11|6|6|17 11|6|6|17 14.0 (28/02) 11.3 (34/03) 10.0 (40/04) 9.2 (46/05) 8.6 (52/06) 8.2 (58/07) 8.0 (64/08) 7.7 (70/09) 5.6 (76/10)
-9 10|6|6|17 10|6|6|17 13.5 (27/02) 11.0 (33/03) 9.7 (39/04) 9.0 (45/05) 8.5 (51/06) 8.1 (57/07) 7.8 (63/08) 7.6 (69/09) 7.5 (75/10)
-5 10|6|6|16 10|6|6|16 13.0 (26/02) 10.6 (32/03) 9.5 (38/04) 8.8 (44/05) 8.3 (50/06) 8.0 (56/07) 7.7 (62/08) 7.5 (68/09) 7.4 (74/10)
0 09|5|5|15 09|5|5|15 12.0 (24/02) 9.6 (29/03) 8.5 (34/04) 7.8 (39/05) 7.3 (44/06) 7.0 (49/07) 6.7 (54/08) 6.5 (59/09) 6.4 (64/10)
8 09|5|5|14 09|5|5|14 11.5 (23/02) 9.3 (28/03) 8.2 (33/04) 7.6 (38/05) 7.1 (43/06) 6.8 (48/07) 6.6 (53/08) 6.4 (58/09) 6.3 (63/10)
12 09|4|5|14 09|4|5|14 11.5 (23/02) 9.0 (27/03) 8.0 (32/04) 7.4 (37/05) 7.0 (42/06) 6.7 (47/07) 6.5 (52/08) 6.3 (57/09) 6.2 (62/10)
13 08|5|5|14 08|5|5|14 11.0 (22/02) 9.0 (27/03) 8.0 (32/04) 7.4 (37/05) 7.0 (42/06) 6.7 (47/07) 6.5 (52/08) 6.3 (57/09) 6.2 (62/10)
15 08|5|5|13 08|5|5|13 10.5 (21/02) 8.6 (26/03) 7.7 (31/04) 7.2 (36/05) 6.8 (41/06) 6.5 (46/07) 6.3 (51/08) 6.2 (56/09) 6.1 (61/10)
21 08|5|5|13 08|5|4|13 10.5 (21/02) 8.6 (26/03) 7.5 (30/04) 7.0 (35/05) 6.5 (39/06) 6.2 (44/07) 6.0 (48/08) 5.8 (53/09) 5.7 (57/10)
24 08|5|5|12 08|5|4|12 10.0 (20/02) 8.3 (25/03) 7.2 (29/04) 6.8 (34/05) 6.3 (38/06) 6.1 (43/07) 5.8 (47/08) 5.7 (52/09) 5.6 (56/10)
25 08|4|4|11 08|4|4|11 9.5 (19/02) 7.6 (23/03) 9.0 (27/03) 7.7 (31/04) 7.0 (35/05) 6.5 (39/06) 7.1 (43/06) 6.7 (47/07) 6.3 (51/08)
29 07|4|4|12 07|4|4|12 9.5 (19/02) 7.6 (23/03) 9.0 (27/03) 7.7 (31/04) 7.0 (35/05) 6.5 (39/06) 7.1 (43/06) 6.7 (47/07) 6.3 (51/08)
34 07|4|4|11 07|4|4|11 9.0 (18/02) 7.3 (22/03) 8.6 (26/03) 7.5 (30/04) 6.8 (34/05) 6.3 (38/06) 7.0 (42/06) 6.5 (46/07) 6.2 (50/08)
43 07|3|4|11 07|3|4|11 9.0 (18/02) 7.0 (21/03) 8.3 (25/03) 7.2 (29/04) 6.6 (33/05) 6.1 (37/06) 6.8 (41/06) 6.4 (45/07) 6.1 (49/08)
46 07|3|4|10 07|3|4|10 8.5 (17/02) 6.6 (20/03) 8.0 (24/03) 7.0 (28/04) 6.4 (32/05) 6.0 (36/06) 6.6 (40/06) 6.2 (44/07) 6.0 (48/08)
50 06|4|4|10 06|4|3|10 8.0 (16/02) 6.6 (20/03) 7.6 (23/03) 6.7 (27/04) 6.0 (30/05) 5.6 (34/06) 6.1 (37/06) 5.8 (41/07) 5.5 (44/08)
61 06|4|4|09 06|4|3|09 7.5 (15/02) 6.3 (19/03) 7.3 (22/03) 6.5 (26/04) 5.8 (29/05) 5.5 (33/06) 6.0 (36/06) 5.7 (40/07) 5.3 (43/08)
67 06|4|4|08 06|3|3|08 7.0 (14/02) 5.6 (17/03) 6.6 (20/03) 5.7 (23/04) 5.2 (26/05) 5.8 (29/05) 5.3 (32/06) 5.8 (35/06) 5.4 (38/07)

[9]

Multiple targets

The table below lists the average frames per bolt when firing continuously at multiple targets which are spread out so that no bolt has to pass through another target (thus even with two targets there's an interval during which the 4 frame next delay expires).


EIAS Frames/Attack Arrows
Server Client 2 3 4 5 6 7 8 9 10
-10 11|6|6|17 11|6|6|17 14.0 (28/02) 11.3 (34/03) 10.0 (40/04) 9.2 (46/05) 8.6 (52/06) 8.2 (58/07) 8.0 (64/08) 7.7 (70/09) 7.6 (76/10)
-9 10|6|6|17 10|6|6|17 13.5 (27/02) 11.0 (33/03) 9.7 (39/04) 9.0 (45/05) 8.5 (51/06) 8.1 (57/07) 7.8 (63/08) 7.6 (69/09) 7.5 (75/10)
-5 10|6|6|16 10|6|6|16 13.0 (26/02) 10.6 (32/03) 9.5 (38/04) 8.8 (44/05) 8.3 (50/06) 8.0 (56/07) 7.7 (62/08) 7.5 (68/09) 7.4 (74/10)
0 09|5|5|15 09|5|5|15 12.0 (24/02) 9.6 (29/03) 8.5 (34/04) 7.8 (39/05) 7.3 (44/06) 7.0 (49/07) 6.7 (54/08) 6.5 (59/09) 6.4 (64/10)
8 09|5|5|14 09|5|5|14 11.5 (23/02) 9.3 (28/03) 8.2 (33/04) 7.6 (38/05) 7.1 (43/06) 6.8 (48/07) 6.6 (53/08) 6.4 (58/09) 6.3 (63/10)
12 09|4|5|14 09|4|5|14 11.5 (23/02) 9.0 (27/03) 8.0 (32/04) 7.4 (37/05) 7.0 (42/06) 6.7 (47/07) 6.5 (52/08) 6.3 (57/09) 6.2 (62/10)
13 08|5|5|14 08|5|5|14 11.0 (22/02) 9.0 (27/03) 8.0 (32/04) 7.4 (37/05) 7.0 (42/06) 6.7 (47/07) 6.5 (52/08) 6.3 (57/09) 6.2 (62/10)
15 08|5|5|13 08|5|5|13 10.5 (21/02) 8.6 (26/03) 7.7 (31/04) 7.2 (36/05) 6.8 (41/06) 6.5 (46/07) 6.3 (51/08) 6.2 (56/09) 6.1 (61/10)
21 08|5|5|13 08|5|4|13 10.5 (21/02) 8.6 (26/03) 7.5 (30/04) 7.0 (35/05) 6.5 (39/06) 6.2 (44/07) 6.0 (48/08) 5.8 (53/09) 5.7 (57/10)
24 08|5|5|12 08|5|4|12 10.0 (20/02) 8.3 (25/03) 7.2 (29/04) 6.8 (34/05) 6.3 (38/06) 6.1 (43/07) 5.8 (47/08) 5.7 (52/09) 5.6 (56/10)
25 08|4|4|11 08|4|4|11 9.5 (19/02) 7.6 (23/03) 6.7 (27/04) 6.2 (31/05) 5.8 (35/06) 5.5 (39/07) 5.3 (43/08) 5.2 (47/09) 5.1 (51/10)
29 07|4|4|12 07|4|4|12 9.5 (19/02) 7.6 (23/03) 6.7 (27/04) 6.2 (31/05) 5.8 (35/06) 5.5 (39/07) 5.3 (43/08) 5.2 (47/09) 5.1 (51/10)
34 07|4|4|11 07|4|4|11 9.0 (18/02) 7.3 (22/03) 6.5 (26/04) 6.0 (30/05) 5.6 (34/06) 5.4 (38/07) 5.2 (42/08) 5.1 (46/09) 5.0 (50/10)
43 07|3|4|11 07|3|4|11 9.0 (18/02) 7.0 (21/03) 6.2 (25/04) 5.8 (29/05) 5.5 (33/06) 5.2 (37/07) 5.1 (41/08) 5.0 (45/09) 4.9 (49/10)
46 07|3|4|10 07|3|4|10 8.5 (17/02) 6.6 (20/03) 6.0 (24/04) 5.6 (28/05) 5.3 (32/06) 5.1 (36/07) 5.0 (40/08) 4.8 (44/09) 4.8 (48/10)
50 06|4|4|10 06|4|3|10 8.0 (16/02) 6.6 (20/03) 5.7 (23/04) 5.4 (27/05) 5.0 (30/06) 4.8 (34/07) 4.6 (37/08) 4.5 (41/09) 4.4 (44/10)
61 06|4|4|09 06|4|3|09 7.5 (15/02) 6.3 (19/03) 5.5 (22/04) 5.2 (26/05) 4.8 (29/06) 4.7 (33/07) 4.5 (36/08) 4.4 (40/09) 4.3 (43/10)
67 06|4|4|08 06|3|3|08 7.0 (14/02) 5.6 (17/03) 5.0 (20/04) 4.6 (23/05) 4.3 (26/06) 4.8 (29/06) 4.5 (32/07) 4.3 (35/08) 4.2 (38/09)


Throw

1HT

This table also applies to Poison Javelin, Plague Javelin, Lightning Bolt and Lightning Fury.


Frames EIAS IAS
20 10 0 -10
20 -23
19 -19 2
18 -15 6
17 -10 11
16 -5 18 6
15 0 24 11
14 8 37 22 9
13 15 50 32 18 6
12 24 70 48 30 16
11 34 99 70 48 30
10 46 147 105 75 52
9 61 250 174 125 89


Reference
Retrieved from "https://www.theamazonbasin.com/wiki/index.php?title=Amazon_attack_speed&oldid=62385"