Q&A Mechanics & Techniques Discussion

[Fox Is Openly Deceptive]

So fun fact, but you know how I mentioned that MK's D-smash starts charging on the first frame? I noted this earlier because it means that he is always going to be at least one frame slower at hitting both hits of D-smash. In other words, while MK's D-smash is listed as hitting on frame 4 in front and frame 9 behind, it is physically impossible to do this (except maybe via buffering?), and at best you'll hit on frame 5 in front and frame 10 behind because you must endure at least one frame of charge every time you use the move.

Edit: Yeah I just checked, it's as I thought, you do have to buffer it in order to avoid the one frame of charge and get it to hit on frame 4 and 9.

 

[Kofu]

Oh my Bionis yes! Thank you Detecitive, I used an input of any kind for the footstool jumpsquat & saw that it shows the input's frame 1 of said action on frame 8. So yeah, Shulk's footstool jumpsquat is 7 frames. The stuff in the next bit is just about what I noticed with the Charge Release once a smash attack was fully charged.

On the topic of releasing a Smash attack charge, I noticed that the charge window for Pikachu's FSmash was one frame before the hitbox comes out. However, in testing, it seems to take two frames to release the charge (I assume one frame to recognize that you're no longer charging and then it continues with the rest of the animation). I haven't tested this for all characters but I assume that the time it takes between releasing a smash attack charge and the hitbox coming out is one frame plus whatever the number of frames is between the charge window and the actual hitbox.

 

[Vipermoon]

charging a smash attack to maximum

Fox Is Openly Deceptive Thanks. I was only questioning it incase you were talking about something else because damn does it feel like at least 90 frames when you're in the moment. And yes, I know it's 60 frames for charging only, that's why I made sure I worded it the way I did^

 

[Shaya]

Backing Foxy up on this. All of Corrin/Rosa's smashes can be charged for 60 frames. Same as Melee/Brawl's number.

EDIT: Remembered this and finally got around to it, though I used Sheik:

Red lightning percent went ~658%→549% (3ds FD) for 0%→150% with freshness bonus against Sheik, which is around a x1.15 increase in knockback.

However, what intrigues me is the rageless kill percent. If freshness bonus is x1.05 damage and Palu jab triggers red lightning on Sheik at 664% in training, I would've expected 645~646% to be where Sheik shows red lightning outside of training, not 658%. Has it been confirmed that the freshness bonus in Smash 4 is still x1.05?

edit2: testing with Rosa's pummel, it seems to be. Does 3% instead of 2% on every 10th pummel. Unsure why the kill percents don't line up then.

A move which deals literal 10% more knockback [i.e. 50 to 55] won't be sending a character 10% further.
So while damage scales near linearly with a move's knockback (but not as linearly as we know rage does), how knockback interacts with an an engine of constant opposing forces won't be "linear".

 

[Lavani]

A move which deals literal 10% more knockback [i.e. 50 to 55] won't be sending a character 10% further.
So while damage scales near linearly with a move's knockback (but not as linearly as we know rage does), how knockback interacts with an an engine of constant opposing forces won't be "linear".

I'm aware of that. A 3% move with 30b/60g should do around 193 units of KB on Sheik at 664%. A 3.15% move with the same knockback numbers and target would be expected to break 193 units of KB at 645%, but that's not the case where freshness bonus is concerned.

 

[Shaya]

That is interesting then.
There very would could be a rounding down (or up) somewhere here.

 

[TreeFire80]

Hey so I was wondering if there's some other resource or thread around here that includes the types of frame data that's not on kuroganehammer.com. Most notably I was wondering what frame a smash attack comes out on after letting go of A, the FAF on throws including both grab releases, and item throw frame data including picking up an item (particularly for characters that spawn items). Is there somewhere you guys can point me to? Does such a data set even exist? If not, is anyone working on it or is there some readily available way to do it on my own without a capture card?

 

[Jams.]

How much additional hitlag do you incur from hitting multiple opponents? Does this depend on when you hit each opponent and does the hitlag overlap (ie. hitting 2 opponents at the same time results in no extra hitlag)? I feel like this is significant in cases like aerial safety versus Rosalina, where you're likely to hit Luma in addition to her shield; I believe Dabuz was able to shieldgrab a lot of ZeRo's fairs because of the additional hitlag incurred from hitting Luma.

Also, as a corollary, do aerials incur any additional frames of lag if they clank?

Hey so I was wondering if there's some other resource or thread around here that includes the types of frame data that's not on kuroganehammer.com. Most notably I was wondering what frame a smash attack comes out on after letting go of A, the FAF on throws including both grab releases, and item throw frame data including picking up an item (particularly for characters that spawn items). Is there somewhere you guys can point me to? Does such a data set even exist? If not, is anyone working on it or is there some readily available way to do it on my own without a capture card?

Sixriver has data for everything you would like I believe, though it's in Japanese. Pikabunz made an amazing translation of some of the miscellaneous data on that website including data you want like item tosses and smash attack releases. Please keep in mind sixriver uses total frames instead of FAF for Kuroganehammer, so actual values will be a frame off (total frames = FAF - 1). I'm not sure if a more comprehensive translation exists for the data on that website.

 

[Fox Is Openly Deceptive]

is there some readily available way to do it on my own without a capture card?

http://smashboards.com/threads/foxys-guide-to-testing-and-labbing.417822/

 

[KuroganeHammer]

How much additional hitlag do you incur from hitting multiple opponents? Does this depend on when you hit each opponent and does the hitlag overlap (ie. hitting 2 opponents at the same time results in no extra hitlag)? I feel like this is significant in cases like aerial safety versus Rosalina, where you're likely to hit Luma in addition to her shield; I believe Dabuz was able to shieldgrab a lot of ZeRo's fairs because of the additional hitlag incurred from hitting Luma.

Also, as a corollary, do aerials incur any additional frames of lag if they clank?



Sixriver has data for everything you would like I believe, though it's in Japanese. Pikabunz made an amazing translation of some of the miscellaneous data on that website including data you want like item tosses and smash attack releases. Please keep in mind sixriver uses total frames instead of FAF for Kuroganehammer, so actual values will be a frame off (total frames = FAF - 1). I'm not sure if a more comprehensive translation exists for the data on that website.

Also you can directly ask me here or on tweeeeeeeter. Usually I know/can find the answer for most frame data things.

 

[TheReflexWonder]

Does the Sakurai angle have the same linear growth for grounded opponents as in Brawl? More importantly, does it cap at a 37-degree angle at high knockback values in Smash 4?

 

[Lavani]

Do we know the exact difference between 365° and 366° yet? Both are autolinks, but surely there's a difference or else there wouldn't be a point in having two angles for it.

Does the Sakurai angle have the same linear growth for grounded opponents as in Brawl? More importantly, does it cap at a 37-degree angle at high knockback values in Smash 4?

Growth looks the same. Starts ramping up after 60KB, seems to plateau around 88KB. Eyeballing the actual angle is kind of hard but 37° looks plausible.

 

[Masonomace]

I'm having a slight problem with calculating hitlag aka freeze-frames, but maybe I just need to ask a simple question so that it would make sense if the answer was what I'm thinking it might be.

Question(s):
Does the first frame of a hitbox also count as the first frame of hitlag, or does hitlag start to occur after the attack's hitbox connecting? (Apologies in advance if this is a silly question.)

Has the hitlag modifier for the first hit of Shulk's Fsmash changed?? The pastebin here shows it's 1.5x, but since this pastebin is only 1.0.4, maybe it was reduced over the course of patches. . .

Spoiler: Math stuff

Vanilla Shulk's first hit of Fsmash deals 5.5%. Looking at A_kae's post about hitlag, I punch in numbers:

((5.5/2.6) +5)*1.5*1 = 10.6730769225 = 10 freeze-frames. However, when I count frames in Training Mode, I get 9 freeze-frames??

Smash Shulk's first hit of Fsmash deals 2.75%. So I punch in numbers:

((2.75/2.6) +5)*1.5*1 = 9.086538462 = 9 freeze-frames. However, when I count frames in Training Mode, I get 8 freeze-frames??

 

[Sonicninja115]

I have heard many times that Ness, Lucas and Mewtwo's airdodges have something special about them. Is this true and does anyone know what it is?

 

[Masonomace]

I have heard many times that Ness, Lucas and Mewtwo's airdodges have something special about them. Is this true and does anyone know what it is?

When you simultaneously doublejump & then airdodge, the doublejump itself has a height difference. For Ness, he goes higher, but Lucas & Mewtwo go less high.

 

[Yikarur]

Ness, Lucas and Mewtu get control over their momentum if they airdodge after hitstun. Thats unique to them.

 

[DanGR]

Question(s):
Does the first frame of a hitbox also count as the first frame of hitlag, or does hitlag start to occur after the attack's hitbox connecting?

Hitlag only occurs when a hitbox connects, and the hitlag begins on the same frame of contact.

For example, for an attack that has 5 frames of hitlag:
It hits a shield on frame 6.
Hitlag on frames 6-10
6, 7, 8, 9, 10

 

[HimaBook]

However, what intrigues me is the rageless kill percent. If freshness bonus is x1.05 damage and Palu jab triggers red lightning on Sheik at 664% in training, I would've expected 645~646% to be where Sheik shows red lightning outside of training, not 658%. Has it been confirmed that the freshness bonus in Smash 4 is still x1.05?

Sorry for long sentence, but it's needed.
To put it briefly, freshness bonus is also x1.05, but KB formula is a little differet.The following are the details.

I guess you assigned “(Base Dmg.)*SMN ” to target damage of quoted formula (below) ,when you calculate KB .
(SMN: Effect of stale-move negation . e.g. If fresh ,1.05 . Base Dmg: Damage without SMN (Same to case of SMN=1).)

Knockback = ((((((t+d)/10+(((t+d)*d)/20))*(200/(w+100))*1.4)+18)*(g/100))+b)

t = Target Damage
d = Attack Damage
w = Target Weight
g = Knockback Growth
s = Weight Based Knockback
b = Base Knockback

If so, it’s not correct. In Inose’s research article (Written in Japanese) , formula is a little different if SMN is considered.To calculate correctly, the formula has to be changed as this fig.

So the formula should be written as below using Base Damage(BD) and SMN.

Knockback
=((((((t+BD*SMN)/10+(((t+BD*SMN)*BD*(1-(1-SMN)*0.3))/20))*1.4* (200/(w+100)))+18)*(g/100))+b)

t = Target Damage
BD = Base Damage of attack
w = Target Weight
g = Knockback Growth
b = Base Knockback
SMN=Stale Move Negation

This formula makes your result (quoted first) clear. Below is calculation.

Palutena’s jab: BD=3, g=60, b=30 , Sheik’s weight: w=84

(1)In training: SMN=1, t=664 (red lighting % in your result)

・・・・calculated KB≒193.05​

(2)In the case of freshness bonus: SMN=1.05, t=658 (red lighting % in your result)

・・・・calculated KB≒193.07​

As can be seen, these 2 cases have nearly equal KB by this formula. So freshness bonus is also x1.05 in smash4 , but formula is a little different.

 

[nannerham]

Hey guys here's a nifty kill confirm lucario has had hidden within him (I'm still yet to get input from other lucario mains whether this is new or not)

Here some cool things about the kill confirm (ASC-> reverse fair-> bair) :

1. Its a true combo
2. It acts as a DI trap for when your opponent tries to get out of ASC-JCUS (aura sphere charge jump cancel up smash)
3. It actually can kill earlier than ASC JCUS
4. While the video makes it seem the window for this is tight you actually have a pretty big window of % for when this actually works (20% for fast fallers and 15% for floaty characters)

This is a true combo on every character except: ness, lucas, kirby, g&w, and jigglypuff but it turns into a 50/50 kill set up on them. This combo is at its deadliest on smashville and can net kills as early as 60% and of course aura and rage make %'s vary. Player 2 in the video was using optimal DI (which is up and away) and DI'ing any other way is deadly, if you DI down, away or both its easier to get hit with bair and you end up losing your stock at earlier %'s, if you DI in you're in perfect position to get hit by uair and thats also deadly. If you have questions then ask away thanks for reading this.

 

[HimaBook]

Except SMN ,I want to refer about rage effect formula in same Inose’s article (Japanese).

He said rage effect formula is (1+(x-35)*0.0013) .

If 35≤(own dmg.)≤150, x=(own dmg.)
If (own dmg.)<35, x=35
If (own dmg.)>150, x=150​

I emphasize It can be transformed into (0.0013x+0.9545).
It is very similar to Lavani’s formula. (0.001304x + 0.954348)

My current understanding of rage, where x denotes your own percent and y is the multiplier on final knockback,

if x ≤ 35, y = 1
if 36 ≤ x ≤ 150, y = 0.001304x + .954348
if x ≥ 150, y = 1.15

Lavani focused on rage effect of 150% to be x1.15, but Inose fixed rage effect of 35% to be x1.
It appears only this point is difference. Because two research indicate similar formula, these are believable.

 

[Lavani]

Thank you very, very much, HimaBook ! Glad to have these things cleared up~

Incidentally, this all got me thinking I should be looking at more Japanese resources for Smash, and I stumbled on the answer to my latest question here:

Do we know the exact difference between 365° and 366° yet? Both are autolinks, but surely there's a difference or else there wouldn't be a point in having two angles for it.

If I'm interpreting it right (pretty sure I am, it looks accurate in-game):

365° - Knockback angle matches the user's movement.
366° - Upward knockback on grounded opponents (85°?), autolink on airborne. Pulls opponents toward the user's center in addition to matching movement.
367° - Pulls opponents toward the user's center in addition to matching movement.

Also not something I'd ever stopped to think about, but it seems autolink hitstun is calculated from what the move's non-autolink knockback would be. Explains why Greninja uair starts putting Sheik in tumble around 67% even if she isn't really launched anywhere.

 

[HimaBook]

Lavani , do you read Japanese? Or are you Japanese?Either way, I have to write what you should note to read Inose’s article for avoiding confusion.Please note following.

(1)In my reply quoted below, I transformed original formula to similar form of KB formula posted here.

Knockback
=((((((t+BD*SMN)/10+(((t+BD*SMN)*BD*(1-(1-SMN)*0.3))/20))*1.4* (200/(w+100)))+18)*(g/100))+b)

t = Target Damage
BD = Base Damage of attack
w = Target Weight
g = Knockback Growth
b = Base Knockback
SMN=Stale Move Negation

It's for ease to understand. But it’s equal to original formula.
Originally, Inose's KB formula is

Knockback
=((0.07*(BD*(1-(1-SMN)*0.3)+2)*(t+(BD*SMN))/(w’/100)+18)*g’+b)*(1+(x-35)*0.0013)

(2)“weight(重さ)” in original formula is not equal to weight we can see Kurogane Hammer (So I wrote as w’ , not w). Relation of these is w’=100+(w-100)/2

(3)In original formula,Knockback growth is not　in % ,but based on 1(So I wrote as g’ , not g).Relation is g'=g/100

(4)Original formula includes rage effect as (1+(x-35)*0.0013),which I removed in my reply.

 

[Meshima]

Here is the formula about IASA frames from Hitlag and Shield Stun in 1.1.4.

Hitlag(against Non-Shielded Opponents)

INT((Dmg/2.6+5)*Hitlag Multiplier*Electric Multiplier)
Note; There is 30F limitation.​

Hitlag(against Shielded Opponents)

INT((Dmg/2.6+5)*Shieleded Hitlag Multiplier*Electric Multiplier)
Note; There is 30F limitation.​

1 > Shieleded Hitlag Multiplier

Shieleded Hitlag Multiplier = Hitlag Multiplier​

1 <= Shieleded Hitlag Multiplier <= 1.25

Shieleded Hitlag Multiplier = 1​

1.25 < Shieleded Hitlag Multiplier

Shieleded Hitlag Multiplier = 0.8 * Hitlag Multiplier​

Shield Stun(With Normal Shield, against Non-Projectile)

INT(Dmg/1.72+3)​

Shield Stun(With Power Shield, against Non-Projectile)

INT(Dmg/2.61+3)
Note; There is an error with Roy's full charged NB but couldn't find accurate formula.​

Shield Stun(With Normal Shield, against Projectile)

INT(Dmg/3.44+3)​

Shield Stun(With Power Shield, against Projectile)

INT(Dmg/5.22+3)​

Spreadsheet

 

[KuroganeHammer]

Realistically it should be +2 not +3 because you're calculating shield stun to use in other formulas.

 

[I speak Spanish too]

When you do a RAR aerial, does the turnaroud jump take a frame because of the turnaround animation?
For example, if I do a standing b-air with Corrin the first hit is frame 13.
But If I do a turnaround b-air will it be 14 or still 13?

 

[Megamang]

Hey guys here's a nifty kill confirm lucario has had hidden within him (I'm still yet to get input from other lucario mains whether this is new or not)

Here some cool things about the kill confirm (ASC-> reverse fair-> bair) :

1. Its a true combo
2. It acts as a DI trap for when your opponent tries to get out of ASC-JCUS (aura sphere charge jump cancel up smash)
3. It actually can kill earlier than ASC JCUS
4. While the video makes it seem the window for this is tight you actually have a pretty big window of % for when this actually works (20% for fast fallers and 15% for floaty characters)

This is a true combo on every character except: ness, lucas, kirby, g&w, and jigglypuff but it turns into a 50/50 kill set up on them. This combo is at its deadliest on smashville and can net kills as early as 60% and of course aura and rage make %'s vary. Player 2 in the video was using optimal DI (which is up and away) and DI'ing any other way is deadly, if you DI down, away or both its easier to get hit with bair and you end up losing your stock at earlier %'s, if you DI in you're in perfect position to get hit by uair and thats also deadly. If you have questions then ask away thanks for reading this.

How did you get a game to start with the percentages like that?

Also, sweet find.

 

[nannerham]

How did you get a game to start with the percentages like that?

Also, sweet find.

Switched on handicap it saves a lot of grief with specific rage and aura %'s

 

[HimaBook]

When you do a RAR aerial, does the turnaroud jump take a frame because of the turnaround animation?

RAR requires additional 1F at least. RAR Bair mechanism is following.
run → turnaround → interrupt turnaround with jump → Bair
So you need 1F of turnaround at least (you can interrupt turnaround with jump from 2F). Thus, RAR requires additional 1F.

In other words ,when your character is running to right, left & jump inputs at same time don't cause RAR. The inputs cause jump to left remaining facing to right.
So for RAR ,you have to wait to input jump at least 1F from the frame you input turnaround .

 

[E.Lopez]

Hi! I'm somewhat new to this side of smashboards, and I have a question that is probably already answered somewhere but I don't know where to look, sorry.

Anyway, a mechanic I'm familiar with is being able to kill opponents at earlier percents than normal if I hit them while they are in their smash charge animation.

For example, Samus' down-tilt normally kills Fox around 180% in training mode with no DI, but during a match last night I hit him with down-tilt while he was just starting up his up-smash and he died at 110% (I was at 120% with rage but still that's a big difference from 180%).

How exactly does this mechanic work? Is there a set amount of knockback added to all moves if they land on an opponent who is charging a smash attack?

 

[nannerham]

Hi! I'm somewhat new to this side of smashboards, and I have a question that is probably already answered somewhere but I don't know where to look, sorry.

Anyway, a mechanic I'm familiar with is being able to kill opponents at earlier percents than normal if I hit them while they are in their smash charge animation.

For example, Samus' down-tilt normally kills Fox around 180% in training mode with no DI, but during a match last night I hit him with down-tilt while he was just starting up his up-smash and he died at 110% (I was at 120% with rage but still that's a big difference from 180%).

How exactly does this mechanic work? Is there a set amount of knockback added to all moves if they land on an opponent who is charging a smash attack?

This is a mechanic that's been around since melee, the more you charge a smash attack, the lighter your character becomes, 120 percent rage+charging a smash attack=a much earlier death.

Edit: you actually receive 1.2x the knockback while charging the fsmash, thanks for the correction.

 

[HimaBook]

How exactly does this mechanic work? Is there a set amount of knockback added to all moves if they land on an opponent who is charging a smash attack?

It's not addition of knockback, but multiplication.
In the case of the opponent charging smash ,multiply knockback formula by 1.2 (ref. Inose's article, Japanese).

 

[Vipermoon]

The more you charge? It's supposed to be 1.2x knockback no matter your charge time.

 

[KuroganeHammer]

Ignore it's 5am and I can't read

 

[HimaBook]

Does anyone know relation between angle of moves and vertical/horizontal distance of knockback ? If it was same to physics, vertical and horizontal distance should depend on “knockback*sinθ” and “knockback*cosθ” respectively (θ=angle of moves). But in smash4,it doesn't.

 

[Lavani]

Does anyone know relation between angle of moves and vertical/horizontal distance of knockback ? If it was same to physics, vertical and horizontal distance should depend on “knockback*sinθ” and “knockback*cosθ” respectively (θ=angle of moves). But in smash4,it doesn't.

Gravity increases vertical (but not horizontal) knockback to offset increased falling acceleration, so characters like Fox still suffer "normal" upwards knockback compared to characters like Mario. I'm guessing that's what's messing things up since I haven't actually looked that closely at it.

 

[Xygonn]

Also 80KB. The transition to tumble knockback is what makes spikes start groundbouncing.

I should have paid more attention to this thread. I didn't realize this and figured it out on my own through testing Just searched to see if it was in here. There it is.

Also, I have a formula that is OK at predicting vertical KOs, but not perfect. Somehow fall speed or something fits in there.

Anyway, it's just: Vertical KB / ((2*gravity)^(1/2.5)) > 395 (for battlefield) should give a KO.

For vertical KB I use sin(angle)*raw KB*(1+(gravity-0.075))*5

 

[Ssbm_Jag]

What is the rule for catching items with an aerial?

Will a move only catch items on start-up? Do we know the frame window to when a move will/will not catch an item? Bombs and Banana are relevant items.

 

[Fox Is Openly Deceptive]

What is the rule for catching items with an aerial?

Will a move only catch items on start-up? Do we know the frame window to when a move will/will not catch an item? Bombs and Banana are relevant items.

Interesting question.
Testing method:

Spoiler

To test this I set up a custom stage with a flat base and slightly slanted platform above it that went the length of the base. I then went into training mode and made the cpu a Falco, put Falco below standing on the base, moved player 1 up top, picked a spot and spawned a banana, then moved Falco directly beneath it. I then had Falco do a FH buffered Nair and, using frame skipping techniques, I counted the amount of Nair frames that came out before the banana was caught. By moving player 1 slightly each time before I spawned the banana, I could slowly adjust the amount of Nair frames that came out before the banana was caught until the buffered Nair no longer caught the banana. The furthest it got was to the point that I was catching the banana on frame 5 of the Nair. I then tested it for Falco's other aerials and for airdodge and got the same result. I then did the same thing with Diddy; I wasn't about to test all characters, but I figured I should check to see if there are any differences with an item specialist character, and I got the same results.

There's a 5 frame window. Aerials and airdodge will catch items on frames 1 to 5.

 

[Ssbm_Jag]

Fox Is Openly Deceptive Thanks, if this is truly accurate, and it sounds like it is, this info should be on the SSB Wiki so it's easier to access. I'm not in charge of that though.

 

[FlynnCL]

Just a really quick dumb question I have about shield stun. I'm sorry if this is the wrong place to ask.

Is the formula also affected by the 1.05x freshness bonus and stale move negotiation? Dr. Mario's back-air would do 10 frames of stun instead of 9 if that's the case which would be really nice.