Cyrax Frame Data

[Sprint]

Cyrax Frame Data (updated 25/6/11)

EXE: First active frame of the move
CAN: Duration of the move when chained or cancelled
DUR: Duration of the move when left to complete

FRAME-ADV: Frame advantage on block/hit.
(many moves have a special property that forces this to zero)

CANCELADV: Frame advantage when cancelled into another move or chain.

NOTES: L/M/H. H+ for moves that whiff crouchblock.

MOVE         EXE CAN DUR FRAME-ADV CANCELADV NOTES
                          BLK  HIT  BLK  HIT
FP             9  13  22    0    0  +16  +22  H
> FP          15  24  25    0    0  +11  +25  H
> > FP        17  --  47   -3   xx            M
> BP          18  27  39    0    0  +18  +32  H
> > FP        12  13  42  -10   xx  +19   xx  H
> > FK        10  22  37    0   xx  +15   xx  H+
BP            12  19  29    0    0  +20  +32  H
> FP          15  22  29    0    0  +20  +29  H
> > BP        28  --  55    0   xx            H  (can otg after, blockable)
> BP          20  26  47    0    0            H
> > f+FP      19  --  66    0   xx            M
FK            10  11  36   -6   +5  +19  +30  H
> FK          12  14  36   -4   +4  +18  +26  H+
> > BK        14  --  36   +5    D            H+ 
BK            19  --  49   -3   +5  +18  +26  H
d+FP           6  11  25  -11   +1   +1  +15  H
d+BP          10  --  39   -2   xx            H+
d+FK           9  13  24   -8   +1   +3  +12  L
d+BK          13  16  30   +1  +16  +25  +30  L
b+FP          14  --  37   +1    0            T  (9f tech window from 14-22f)
> BP,FP,BP                     +34            T
b+BP          30  --  --   -5  +21            M  (hit.adv. if explosion whiffs)
> explosion   +9  --  64   -2   xx            H  (blk.adv. if 1st hit whiffs)
b+FK          16  --  54  -11   xx            H
b+BK          18  --  52   -7   xx            L
f+BP          29  31  45  +11   ??  +25  +40  H  (blk.adv. if second hit whiffs)
> followup    15  --  51   --   xx            H  (automatic on hit)
u+BK          14  --  46   -5   xx            M
Any Throw     10  --  36    0   xx            T

nj.P           9  --  ??   +3   xx            M  (blk.adv. at lowest height)
nj.K          10  --  ??   +5   xx            M  (blk.adv. at lowest height)
j.P            6  --  ??  +17  +32            M  (adv. at lowest height)
j.K            7  --  ??  +18   xx            H  (blk.adv. at lowest height)

d,f,FP         9  --  31   --   xx            U  (vs air only)
d,b,FP        --  --  46   --   --               (22f before teleport)
b,f,BP        17  --  89  -90   xx            H  (134f duration on block)
b,b,FK        18  --  59  -21  +84            H  (comboable)
d,f,FK        22  44  48   -2   xx            H
> FP+FK       12  --  48   --   xx            U
Any Bomb     114  --  41   --   xx            U  (bomb appears at 13f)

d,f,FP+BL      9  --  34   --   xx            U  (vs air only)
d,b,FP+BL     11  --  46   -8   xx            H  (22f before teleport)
b,f,BP+BL     18  --  64  -90   xx            H  (135f duration on block)
> projectile  +3  --  64  -16   -3            H  (when the above doesn't hit)
b,b,FK+BL     18  --  59  -21  +83            H  (comboable, drains 1 bar)
d,f,FK+BL      9  --  --   --   --            L
> hit 2      +16  ??  47   -1   xx            H
> FP+FK       12  --  48   --   xx            U
Sticky Bomb   18  --  52   -7   xx            H

XRay          15  --  60   +2   xx            H+

Mostly accurate. It's hard to measure things with this stupid game engine.
- Many moves have a special flag that makes them neutral unless you cancel into another move.
- Links are always blockable unless the opponent is put into a comboable state (Net, Launchers), or you hit them with something unblockable.
- No active frames because it's impossible to tell where they are without a lot of tedious work.

Calculate whiff advantage by subtracting move 2 duration from move 1 cancel advantage:
- f+BP xx Bomb on hit = 40 - 41 = -1f

Calculate gaps by subtracting move 1 cancel advantage from move 2 startup (EXE - 1)
- FP,BP on block = (18 - 1) - 16 = 1f gap.
- A 0f gap allows switching between standing and crouching.
- A 1f gap allows Player B to reversal.
- A 2f gap allows Player A to reversal.
"Player A" refers to whoever the game assigns to win trades for a particular match. Win trades, lose to frametraps.

Updated for latest patch (as of June 25). Major changes are:
- Command grab hitstun reduced by 30f
- Command grab tech window increased to 9f from 2f
- f+BP hitstun reduced by 17f
- Net damage reset no longer works.
- Net cooldown time significantly reduced.
- Net allows the opponent to tech throws.
- Various minor changes that I think were typos on my part.

 

[Juggs]

Very nice man! Thread stuck. rink:

 

[FiNAL-LiFe-G]

wat is > > FP? is that forward, forward, Square?

 

[Sporko]

No, that indicates it's the next hit in the string.

FP (first hit of the string)
>FP (second hit of hte string, i.e. FP, FP)
>>FP (third hit of the string, i.e. FP, FP, FP)

so the data next to >>FP is the frame data on the total string and the data next to >FP is if you stop the string before it finishes.

 

[Mechacide]

Out of interest, how did you test the hit/block-advantage/disadvantage?

 

[Creepy00]

Would you guys say this frame data is fairly accurate?

 

[Somberness]

Would you guys say this frame data is fairly accurate?

Mostly, although I did not check much.

 

[Seapeople]

Hey Somberness do you know how much advantage Cyrax is at after the opponent falls from an air net or after a second net on the ground?

 

[Playpal]

Hey Somberness do you know how much advantage Cyrax is at after the opponent falls from an air net or after a second net on the ground?

I was in the lab and was experimenting with 21 net, 21 net. According to the frame data here, I was able to get a guaranteed 2, which means double net from the ground is at least +12. If the opponent is in the air though, many variables need to be considered such as gravity of fall and height. Also if the second net hits the opponent while they're too high, you'll be at less advantage than if the opponent was closer to the ground. What really interested me was that during the stagger state of the second net, the opponent was invincible, meaning there's no way to get a "guaranteed" attack; the opponent will always be able to armor out.

 

[Seapeople]

I was in the lab and was experimenting with 21 net, 21 net. According to the frame data here, I was able to get a guaranteed 2, which means double net from the ground is at least +12. If the opponent is in the air though, many variables need to be considered such as gravity of fall and height. Also if the second net hits the opponent while they're too high, you'll be at less advantage than if the opponent was closer to the ground. What really interested me was that during the stagger state of the second net, the opponent was invincible, meaning there's no way to get a "guaranteed" attack; the opponent will always be able to armor out.

That's interesting. Yeah I know what you mean about doing a second net in an air combo. If the net disappears instantly I think you're left at neutral.

Another thing I've found is that opponents can actually X-ray & sometimes attack to escape Cyrax's corner reset. (21~net, jip 121~ex bomb, close bomb, 3~net). I think it's only escapable if you do 3~net too early though

 

[Playpal]

The invincibility would imply that even the guaranteed resets associated with double net are escapable with armor or xray. I'll have to do some more testing, but maybe double net on an airborne opponent is different. This would also explain why sometimes double net resets won't work; the opponent will break out of the net and the bomb will just explode.

 

[NariTuba]

The invincibility would imply that even the guaranteed resets associated with double net are escapable with armor or xray. I'll have to do some more testing, but maybe double net on an airborne opponent is different. This would also explain why sometimes double net resets won't work; the opponent will break out of the net and the bomb will just explode.

Theres 2 types of 2nets (second nets), short duration dissolve and long duration dissolve.

As far as I understand it (and I dont think anyone really does, especially not NRS ) you generally get the first type when you catch someone high in the air (like you pointed out) and you get the second type (long dissolve) when you catch them low to the ground.

The interesting thing is that the invincibility changes sometimes while the stagger does not. Gravity might play a part in this. Double net resets need the long dissolve to keep the opponent near the bomb while staggering. Funny thing is the ublockable properties of the bomb do NOT override 2net invincibility (not in all cases at least) as I found while experimenting with Cyblax. Tony-T found 2net finishers that use the short dissolving net and still grant advantage (see Round 5)

Itd be interesting if Somberness were to get some frame data on the established 2net combos (Tony-Ts and Mustards) to check what we're dealing with here. Im afraid the true interaction between height, gravity, 2net dissolve and advantage might never be fully understood (for the same reason this character is so troublesome to fix). What we CAN do however is continue to find those unique instances where the 2net has advantage (aka 2net combos).

Do share your findings

 

[Playpal]

I'm doubtful that the double net resets we've seen so far are escapable, such as the 75% double net reset by Tony-T we see so much. People must have tested it plenty of times (including Tony-T), so I'm highly doubtful I'm going to be the black sheep that figures out it's escapable. I've done some testing on the dummy and it just doesn't seem escapable. I'm guessing that A) There are more stagger state frames than there are invincibility frames or B) invincibility frames are different when opponent is double netted while airborne. Or it could be C) both. I tested the advantage on this combo: 21 net, jip 21 ex bomb, mid bomb, 123 (mid bomb hits), net. It seems to be roughly neutral. I tested it again, this time with linking 123 into bomb (therefore launching the opponent for a guaranteed reset) and cyrax recovers after the opponent is launched. If it was before the opponent launched, then that would mean the opponent could escape.

tl;dr: my previous theory of being able to escape double net reset is disproved.

 

[NariTuba]

A) There are more stagger state frames than there are invincibility frames or B) invincibility frames are different when opponent is double netted while airborne. Or it could be C) both.

Maybe the invincibility time is different somehow for unblockables vs normals (by accident of course, not by design).


The following corner combos are interesting test subjects:

jps,2,1~net, jps,2,1~Exb, cb, net

depending on the speed (height) of your net wether its escapable or not

jps,1,2,1~net, (pause) 3,3~net

This oddball combo grants advantage on the 2net. If you were to add a jk before the 3,3~net you lose the advantage.

 

[Playpal]

This frame data must be really outdated cuz my tests show that b2 is advantage (explosion must chip).

 

[Seapeople]

This frame data must be really outdated cuz my tests show that b2 is advantage (explosion must chip).

That could be really helpful. But it does say "block adv if 1st hit whiffs."

Do you know if the amount of advantage after a command grab is consistent? Sometimes I get a free jump-in, but sometimes they can jump out. I think other people have noticed this problem too

 

[Playpal]

That could be really helpful. But it does say "block adv if 1st hit whiffs."

Do you know if the amount of advantage after a command grab is consistent? Sometimes I get a free jump-in, but sometimes they can jump out. I think other people have noticed this problem too

I don't understand, B2 chips twice, the bonk on the head and then the explosion. I highly doubt the whiff on the first hit would affect the advantage of the explosion. Not to mention I don't even know how the first hit could whiff anyways if the second hit chips, hell I think a lot of this frame data is strange and likely inaccurate. Anyways I highly doubt command grab randomly changes advantage. Given how it's supposedly +34, I'm guessing the window for a safe jump is really tight. Most safe jumps are usually at least +37.

 

[Seapeople]

I don't understand, B2 chips twice, the bonk on the head and then the explosion. I highly doubt the whiff on the first hit would affect the advantage of the explosion. Not to mention I don't even know how the first hit could whiff anyways if the second hit chips, hell I think a lot of this frame data is strange and likely inaccurate. Anyways I highly doubt command grab randomly changes advantage. Given how it's supposedly +34, I'm guessing the window for a safe jump is really tight. Most safe jumps are usually at least +37.

That makes sense. I've pretty much stopped jumping in afterward altogether. And I agree, the frame data probably needs to be redone eventually

 

[Playpal]

Would like to add that ex buzz saw as a projectile is advantage anywhere besides close range. You can spam it full screen almost like kabal's iagb. At close range however, it works just like regular buzz saw and is disadvantage.

Also, you can recover fast enough from uppercut to follow up with d1 in the corner. Would be great if it was actually possible to do d1 net.

 

[NariTuba]

I have a request for Somberness if possible at some point.

It would be useful to know how much hit advantage the following splats provide before an opponent can perform a wakeup attack:

u4
njp splat (for example 3 consecutive njps)
saw
last hit of 2,1,2

The idea is to plan bomb setups around them.

Thank you!

 

[Seapeople]

I have a request for Somberness if possible at some point.

It would be useful to know how much hit advantage the following splats provide before an opponent can perform a wakeup attack:

u4
njp splat (for example 3 consecutive njps)
saw
last hit of 2,1,2

The idea is to plan bomb setups around them.

Thank you!

Ohhh I never thought about using u4. That's a pretty fast, pretty safe overhead so it could be great in the corner

I use the 212 set up all the time. I don't know the specific frame data before a wakeup, but it's really delayed. And the other player might get thrown off because they expect the 334 followup

 

[Somberness]

I have a request for Somberness if possible at some point.

It would be useful to know how much hit advantage the following splats provide before an opponent can perform a wakeup attack:

u4
njp splat (for example 3 consecutive njps)
saw
last hit of 2,1,2

The idea is to plan bomb setups around them.

Thank you!

It's been a long ride but I will have human Cyrax done pretty soon. Of the stuff I have found out... 212 is +44 always since you can't tech roll it. U4 is +44 with no tech roll and +52 with. Buzzsaw is +28. This is all for human, I have noticed differences between Sprint's data but that could just be errors/updates. I still don't get the duration calculation for moves in the data here.

 

[NariTuba]

It's been a long ride but I will have human Cyrax done pretty soon. Of the stuff I have found out... 212 is +44 always since you can't tech roll it. U4 is +44 with no tech roll and +52 with. Buzzsaw is +28. This is all for human, I have noticed differences between Sprint's data but that could just be errors/updates. I still don't get the duration calculation for moves in the data here.

Thank you! awesome stuff man...

so here we have it, u4 gives you as good a corner bomb setup as 212 and much better than saw.

Now up there in the data it says that the duration for any bomb is 44f (and the bomb appears at 13f). So that means that:

u4 + bomb / 212 + bomb = neutral if just framed

saw + bomb = -16

As you say, I dont really understand the duration either. You think the above might be right? (Assuming robo & human have the same duration for bombs of course, which might or might not be true)

Honestly, if you can review the robo data at some point it would be great... take all the time you need of course, we all understand its not easy to get and we really appreciate your work!

 

[Somberness]

.

 

[Somberness]

3 NJP advantage is about +57, +65 if tech rolled.

Spoiler

Execution - frame number when a move becomes active, startup is execution number minus 1.
1 - 9
1,1 - 15
1,1,1 - 17
1,2 - 18
1,2,1 - 12
1,2,3 - 10
2 - 12
2,1 - 15
2,1,2 - 28
2,2 - 20
2,2,F+1 - 19
3 - 10
3,3 - 12
3,3,4 - 14
4 - 19
D+1 - 6
D+2 - 10
D+3 - 9
D+4 - 13
B+1 - 14
B+2 - 30
B+3 - 16
B+4 - 18
F+2 - 29
U+4 - 14
Throw - 10
Bomb - 114 (12 startup)
Bomb (Costume 2) - 113 (11 startup)
Net - 18
Net (full screen) - 61
Net (Costume 2, full screen) - 63
Teleport - 9 startup frames until 13 frames of invincibility
Teleport (Costume 2) - 22 frames of invincibility after switching sides
Buzzsaw - 17
Reverse Kick - 22
Anti-Air - 9
Sticky Bomb - 18
Sticky Bomb (Costume 2) - ~18 (16 startup, the bomb starts inside him)
Sticky Bomb (full screen) - 57
Sticky Bomb (Costume 2, full screen) - 62
Electro Net - 18
Electro Net (full screen) - 61
Electro Net (Costume 2, full screen) - 63
Bangport - 11
Bangport (Costume 2) - 24
Saw Blade - 17
Saw Blade (full screen) - 42
Donkey Kick - 9
Power Anti-Air - 9
Cyberdriver - 15

Block Advantage - number of frames that show how fast you and your opponent recover after the move is blocked.
1 - 0
1,1 - 0
1,1,1 - -3
1,2 - 0
1,2,1 - -10
1,2,3 - 0
2 - 0
2,1 - 0
2,1,2 - 0
2,2 - 0
2,2,F+1 - -20
3 - -6
3,3 - -4
3,3,4 - +5
4 - -3
D+1 - -13
D+2 - -2
D+3 - -9
D+4 - 0
B+2 - +2
B+3 - -11
B+3 (Costume 2) - -7
B+4 - -7
F+2 - +11
U+4 - -5
Net - -21
Buzzsaw - -10
Reverse Kick - -2
Sticky Bomb - -7
Electro Net - -21
Bangport - -8
Bangport (Costume 2) - +4
Saw Blade - -10
Donkey Kick - -2
Cyberdriver - +2
Cyberdriver (Costume 2) - -4

Hit Advantage - number of frames that show how fast you and your opponent recover after the move hits. If there is two numbers, the second one is if your opponent tech rolls.
1 - 0
1,1 - 0
1,1,1 - +22, -1
1,2 - 0
1,2,1 - +45, +28
1,2,3 - +32, +30
2 - 0
2,1 - 0
2,1,2 - +44
2,2 - 0
2,2,F+1 - +5, +3
3 - +5
3,3 - +4
3,3,4 - +15, +19
4 - +5
D+1 - +1
D+1 (vs crouch) - -1
D+2 - +33, +26
D+3 - +1
D+3 (vs crouch) - +8
D+4 - +16
D+4 (vs crouch) - +9
B+1 - 0
B+1,2,1,2 - +34
B+2 - +80, +69
B+3 - +37, +20
B+4 - +17, +9
F+2 - +9, +11
U+4 - +44, +52
Forward Throw - +7
Backward Throw - +7
Bomb - +174, +162
Bomb (Costume 2) - +165, +153
Net - +84
Buzzsaw - +28
Reverse Kick - +46, +31
Ragdoll - +9
Anti-Air - ~+25, ~-5
Sticky Bomb - +95, +84
Sticky Bomb (Costume 2) - ~+97, ~+86
Electro Net - +83
Bangport - +45, +28
Bangport (Costume 2) - +57, +40
Saw Blade - +61, +43
Donkey Kick - +46, +31
Ragdolls - +9
Power Anti-Air - ~+25, ~-5
Cyberdriver - +66

Duration - number of frames it takes to recover after using a move.
1 - 22
1,1 - 38
1,1,1 - 84
1,2 - 52
1,2,1 - 82
1,2,3 - 77
2 - 29
2,1 - 48
2,1,2 - 96
2,2 - 66
2,2,F+1 - 111
3 - 36
3,3 - 47
3,3,4 - 61
4 - 49
D+1 - 25
D+2 - 47
D+3 - 24
D+4 - 30
B+1 - 37
B+2 - 64
B+3 - 54
B+4 - 52
F+2 - 45
U+4 - 46
Throw - 36
Bomb - 41
Bomb (Costume 2) - 49
Net - 59
Net (failed) - 38
Teleport - 46 (22 in front, 24 behind)
Buzzsaw - 74
Reverse Kick - 44
Anti-Air - 31
Sticky Bomb - 52
Electro Net - 59
Bangport - 46 (22 in front, 24 behind)
Bangport (Costume 2) - 51 (27 in front, 24 behind)
Saw Blade - 42
Donkey Kick - 47
Power Anti-Air - 34
Cyberdriver - 44
Cyberdriver (Costume 2) - 50

Cancel Advantage (block) - number of frames that show how fast you and your opponent recover after the move is cancelled and is blocked.
1 - +16
1,1 - +11
1,2 - +18
1,2,1 - +19
1,2,3 - +15
2 - +20
2,1 - +20
2,2 - ≥+20
3 - +19
3,3 - +18
4 - +18
D+1 - +1
D+3 - +2
D+4 - +24
F+2 - +25

Cancel Advantage (hit) - number of frames that show how fast you and your opponent recover after the move is cancelled and lands. If there is two numbers, the second one is if your opponent tech rolls.
1 - +22
1,1 - +25
1,2 - +32
1,2,1 - +74, +57
1,2,3 - +47, +45
2 - +32
2,1 - +29
2,2 - ≥+20
3 - +30
3,3 - +26
4 - +26
D+1 - +15
D+1 (vs crouch) - +13
D+3 - +12
D+3 (vs crouch) - +19
D+4 - +30
D+4 (vs crouch) - +23
B+1 - ≥+7
F+2 - +40

Miscellaneous
Breaker - +16, +15