Official Ask Anyone Frame Things Thread

[ant-d]

if you jump and grab on the same frame which one takes priority?

and what about hitting a and b on the same frame?

Z > X/Y (jump)
B > A

 

[DRGN]

What exactly is the difference between the properties hit-stun and shield-stun? They're not the same thing, right? Or are they, but shields deal with it differently -manipulate the values- somehow? Are they simply calculated off of damage like knockback (speaking of which, what is the knockback formula? If we even have one. I realize we don't really have a good unit for distance)?

Also, what moves don't have normal base knockback? I looked around, and I kind of doubt there's a list somewhere, but who knows.

 

[Lovage]

strong bad answer this u **** wizard

if you grab someone that's shielding, do you have to grab their body? or can you grab their shield?
SUB-QUESITON:
would having a tiny shield make you slightly harder to get grabbed?

 

[ant-d]

if you grab someone that's shielding, do you have to grab their body? or can you grab their shield?
SUB-QUESITON:
would having a tiny shield make you slightly harder to get grabbed?

Cannot grab shields. Must hit yellow.

To illustrate, this will not result in a grab:

 

[Strong Badam]

yeah, the shield can only be hit by standard offensive collisions and not grab collisions. grab requires a hurtbox.

What exactly is the difference between the properties hit-stun and shield-stun? They're not the same thing, right? Or are they, but shields deal with it differently -manipulate the values- somehow? Are they simply calculated off of damage like knockback (speaking of which, what is the knockback formula? If we even have one. I realize we don't really have a good unit for distance)?

You're implying that shieldstun you take would scale with knockback, which is not the case. Hitstun and Shieldstun are actually calculated quite differently.
Shieldstun is calculated with the following formula: ((Damage +4.45)/2.235), rounded down, in frames. Hitstun, however, is calculated by Knockback * 0.4, rounded to the nearest whole frame IIRC. I have an incomplete KB formula and I've been working on figuring out the last pieces of it so I can release it to the public.

Also, what moves don't have normal base knockback? I looked around, and I kind of doubt there's a list somewhere, but who knows.

You'd have to elaborate on what you mean. There are several moves with 0 BKB.

 

[Kal]

I wish I could play Giant Fox in tournament. Up smash spam all day.

 

[kyaputenfarukon07]

Which one has more priority?

Shield Grabbing or Jumping out of Shield

 

[Strong Badam]

This is something you can test yourself. Hold shield, then pause, then hold shield + A + Jump and unpause. See what happens.

 

[kyaputenfarukon07]

Thanks Bro

 

[Bones0]

How soon can you jump out of shield? I'm guessing it's frame 2 (frame 1 the shield comes up), but needed to make sure.

Also, if you start an aerial (Falco's dair for instance) right before you land, do you have the same amount of lag as just a regular empty hop?

 

[ajp_anton]

Pressing jump and shield at the same time will make you shield, and since there is basically no buffering in the game, the jump will never come unless you press it again.
But once you've started shielding, you can jump on any frame starting from the very next one.

There are a few frames in the beginning of an attack animation which can be canceled into a regular empty landing.

 

[Kink-Link5]

What is the frame [dis]advantage of Marth's SHFFL'd fair on block?

 

[Andy Lawliet]

There is no difference between Pivot Grab and regular standing grab, right?

Also when im doing "Pivot" grabs (more like DD JC Grab) and i dont hit the Pivot Frame i often do dash grabs, even while pressing jump like usual

Any specific reason or (more likely) just my messups?

 

[Grim Tuesday]

A pivot x is just x in the opposite direction.

 

[Signia]

What is the frame [dis]advantage of Marth's SHFFL'd fair on block?

Omg someone finally asks a frame data question properly and it's answer would actually be useful.

Most of the questions asked here aren't specific enough or their answers don't tell you anything more than "the length of time is this long." It seems smashers lack the terminology to do ask the right questions, or that terminology isn't established yet. Smashers should really just copy FG terms this time around (lol "friendlies" ... casuals).

I think someone needs to explain some frame data fundamentals. It seems like most of you wouldn't know what do with the data even if you knew it.

One of the few ways to take advantage of specificity of frame data is knowing exactly what beats what in a controlled situation when you know exactly when each player can start doing stuff, like after someone shields an attack or gets hit.

To be clear, the kind of question that's being answered here is: "If I do move and X and it gets shielded, and I do move Y and they do move Z, who will win?" Frame data gives you the tools to answer this question instead of having to memorize things on a case-by-case basis. It also eliminates the chance of confirmation bias causing people think frame traps exist when they don't.

So knowing the shieldstun length of a move isn't particularly important because you also need to know the move's recovery length to see how much sooner or later you act. Your frame advantage will be the move's blockstun minus it's recovery.

Even the frame advantage value isn't worth anything on its own, but now you can use it to see what moves will come out faster.

So if I do a perfectly-low Sheik's Fair on a shield and I jab immediately, will their immediate shield-grab beat it out?

Well, if a perfectly-low Sheik's Fair is -1 on shield, and her jab is i3 (3 frames startup or 3 frames from when you inputted the move to the first active hit frame), that means her jab will hit on frame 4. Since most grabs are i7, the jab will win. This is called a frame-trap.

But what about not perfectly-low Fair? Well, you'll just have to eyeball it. Frames are the smallest unit of time that's represented on the screen, so it's not too hard to estimate how many frames if you compare to things you know for example "well that felt like half the time it takes to grab so I was like 3-4 frames off."

In this situation, jab beats out the grab in the best-case scenario by far. The amount of frames it beats it by is the frames of error you have available. Here, you have 3 frames of error (actually, I don't know what happens when a throw hits on the same frame as an attack so it may be 2 frames, idk).

Actually, it would be more accurate to say you have 3 MORE frames of error than they do, since in the ideal example we assumed they were shield-grabbing on the very first frame they could.

Something to keep in mind though -- when you are doing aerials to land low enough to actually know what you frame advantage is, you're very low and are doing the aerial very late in the SHFFL. This means you have no active hitboxes to protect your jump till then, so you can easily be hit out of the air. For this reason, while doing early aerials is unsafe if you land in their punishing range, they cover more space. So it's a trade-off.



Oh so here is my question: if I'm Doc, is the Up-B cancel glitch-attack a frame trap after just about any aerial or standing attack? I'm invincible on frame 2, right?

But there's no end to these questions. What this community really needs is something like this:

http://8wayrun.com/wiki/yoshimitsu-frame-data-sc5/

Except there are only like 20 moves per character. Columns should list shieldstun, recovery length, active hit frames, frame advantage on hit and block if hit on the first active hit frame and on the last active hit frame.

 

[Tee ay eye]

are there multiple kinds of platform drops? like, are there forward, backward, and neutral platform drops the way there are backwards and neutral jumps?

i see a lot of people platform drop with a certain trajectory, which KINDA leads me to believe there are multiple, but i wasn't sure. when you platform drop in a direction, is it because the platform drop itself is given that directional property, or is it because you do a platform drop directly into aerial control, making it LOOK LIKE it was a directional platform drop?

 

[SUNG666]

How early can I release the ledge after I've grabbed it?

 

[MasterShake]

^ I think you can release on frame 9? Could be 8.

 

[Strong Badam]

frame 9 for everyone but link, frame 5 with link.

working on the marth fair Q

 

[DRGN]

You're implying that shieldstun you take would scale with knockback, which is not the case. Hitstun and Shieldstun are actually calculated quite differently.
Shieldstun is calculated with the following formula: ((Damage +4.45)/2.235), rounded down, in frames. Hitstun, however, is calculated by Knockback * 0.4, rounded to the nearest whole frame IIRC. I have an incomplete KB formula and I've been working on figuring out the last pieces of it so I can release it to the public.

Oops. I didn't mean to suggest that knockback and hitstun/shieldstun used the same formula, only that [perhaps] they too used a formula that is based on damage (I meant the percent damage the opponent/defender is at, but I realize now "damage" in the formulas could instead mean the amount of damage a move deals).

So do you know how SDM and phanna did their data collections on hitlag and shield stun? I tried calculating a few and they didn't add up. The formula I have for hitlag is (d + 2)/3 = f, where d = the damage the move does (I assume), and f = the resulting number of frames of hitlag. Ganon's dash attack does 14%. Plugging that in returns 5.33 frames. phanna has 7 frames written down. Trying shieldstun: ((14 +4.45)/2.235) = 8.255, yet phanna has 15.

You'd have to elaborate on what you mean. There are several moves with 0 BKB.

Most moves use one particular constant for knockback calculations, "Base Knockback", is that right? I was wondering if there is a list of all the moves that do not use it (either they have their own, or have none).

I found this, which looks incomplete.

In this situation, jab beats out the grab in the best-case scenario by far. The amount of frames it beats it by is the frames of error you have available. Here, you have 3 frames of error (actually, I don't know what happens when a throw hits on the same frame as an attack so it may be 2 frames, idk).

Grabboxes beat hitboxes, so it would be 2.

Oh so here is my question: if I'm Doc, is the Up-B cancel glitch-attack a frame trap after just about any aerial or standing attack? I'm invincible on frame 2, right?

Idk about the glitch attack, but with his usual up-b he's invincible on frame 3.

But there's no end to these questions. What this community really needs is something like this:

http://8wayrun.com/wiki/yoshimitsu-frame-data-sc5/

Except there are only like 20 moves per character. Columns should list shieldstun, recovery length, active hit frames, frame advantage on hit and block if hit on the first active hit frame and on the last active hit frame.

I'm thinking (for aerials):

Total number of frames, frames that have a hitbox out, damage the move deals, hitlag, hitstun, shieldstun, IASA, auto-cancel, landing lag (and/or landfallspecial lag)/ l-cancel lag, plus any special notes about the move.

Not sure if I'm missing anything. For ground moves it would be everything except the landing data of course.

Though if any of those includes the percent the opponent/defender is at as a factor, it probably doesn't do much good to include them.

 

[Strong Badam]

Phanna's shieldstun values were found using AR and going frame by frame. They include hitlag. Subtract this: (Damage/3)+3 (again, rounded down) and you should get my values using Phanna's. Some of them are wrong, too.

Knockback is calculated with two KB values. Base Knockback, which is added at the end of the calculation and affects KB the same regardless of opponent % and weight, and Knockback Growth, which is included in some math with the opponent's % and as such "grows" with percent. It's how KB scales with %.
Hitbox damage is also used in KB calculation.

As far as moves that don't use these values, you'd be thinking of Weight Dependent Set Knockback. Some jabs have this, Fox's Shine has it, and so does his drill. There are also moves included on that list that do not use WDSK and simply have 0 KBG, such as Bowser's Up-Throw. Use Masterhand (Pretty sure it's linked in the OP of this thread) to find WDSK moves.

 

[DRGN]

Oh, ok. Interesting. I forgot about that Masterhand program, I'll check it out tomorrow.

KB makes more sense, but I'm still a little unclear about the formulas for hitlag and the stuns. Are you saying to subtract that (Damage/3)+3) from phanna's shieldstun results? So was the formula I had for hitstun (I believe it originally came from Magus420) wrong? Also, to be clear, the "damage" in these isn't the percent damage the opponent is at, but the damage the move does, right?

It's confusing that phanna would write it that way and not even mention that the second number is actually shieldstun PLUS hitlag.

Also, I noticed in your spreadsheets, hitlag is different for the attacker than the defender in some cases?

 

[Grim Tuesday]

Electric attacks and projectiles cause that, I think

 

[ajp_anton]

Is hitlag calculated on the damage it appears to do according to the %-display on the screen, or what the move actually does?
For example, does a 14.5% attack have the same hitlag when the opponent has 5% or 5.5%?

 

[Strong Badam]

I don't think it's rounded before calculation, but I'm not 100% sure. I do know that it uses staled damage.

KB makes more sense, but I'm still a little unclear about the formulas for hitlag and the stuns. Are you saying to subtract that (Damage/3)+3) from phanna's shieldstun results? So was the formula I had for hitstun (I believe it originally came from Magus420) wrong? Also, to be clear, the "damage" in these isn't the percent damage the opponent is at, but the damage the move does, right?

Yes, yes, and correct.

It's confusing that phanna would write it that way and not even mention that the second number is actually shieldstun PLUS hitlag.

It was a different era. SDM's frame data was about all there was. Was certainly before Magus and I started using in-game formulas for things.

Also, I noticed in your spreadsheets, hitlag is different for the attacker than the defender in some cases?

For electric hitlag, the victim suffers 1.5x hitlag and the attacker suffers standard hitlag.
For projectiles, the opponent suffers both hitlag and hitstun and the attacker suffers neither.

 

[ajp_anton]

Meh, will have to do some tests myself then =).

With phantom hits there's also no hitlag for the attacker.

edit: BTW, the hitlag formula is %/3+2 or (%+6)/3.

 

[Blu]

How much stun is there when the reflector is hit by a projectile? (How early can it reflect them?)

How many frames does a turn-around within the reflector last for?

 

[Kink-Link5]

Any luck on Marth's fair data?

Thank you by the way, S B.

 

[Tee ay eye]

1) if you wanted to calculate the magnitude of DI (diminished by not inputting it perpendicularly to the trajectory), would you do....

X*sin(θ)?

Where θ is the angle between the trajectory and the DI input and X is the magnitude of the DI input itself?

p.s. I found that sin^2 post you mentioned a while back:

http://www.smashboards.com/showpost.php?p=12841558&postcount=17995

2) can you input automatic shield DI with the c-stick? i can't test it that easily because of buffering rolls, so i decided i ask here. are shield DI and smash DI controlled by the exact same game mechanic, or are they two separate (but similar) things?

 

[DRGN]

Ok. Thank you for clearing all of that up!

That formula I had may have actually came from Scotu. I'd have to check my sources, which I can't do ATM.

njp_anton, how do you know it is that, versus what StrongBad has? For that matter, how did either of you even discover these? Lol.

 

[Bill_Cosby]

Hello Strang Bod, I was doing battle with a Marth the other day as Samus and noticed that my screw attack performed oddly during two separate attacks.

The first time I UP-B'd from a standing position and Marth, without a shield on, was picked up by it but was shunted from corner to corner of my screw attack as if he was receiving a high knockback hit at high percent (Like how you move backwards a bit when getting knee'd just right at a high percent). I know that the screw attack's hitboxes are in the same corners but this odd movement normally doesn't happen. I think the Marth was at 60-80% at the time.

The second time was also against a Marth and during my recovery Marth went to spike me so I jumped and was going to use my screw attack to intercept him. Near the last hit of the screw attack, however, he was sent downwards as if by a spike.

My question is: Have you heard of Samus' screw attack having any frames where it spikes or has odd directional knockbacks like Pikachu's tail?

 

[Xyzz]

edit: nvm, didn't realize there was another page.

 

[Construct]

Does hitting a shield stale a move?
Does hitting a shy guy/brinstar stuff/green greens block/other stale a move?
Does hitting multiple targets with one move stale it more? Ie I'm marth and I fsmash three people at once, does that stale it three times?

Also bump on the marth f-air question

 

[Bones0]

Does hitting a shield stale a move?
Does hitting a shy guy/brinstar stuff/green greens block/other stale a move?
Does hitting multiple targets with one move stale it more? Ie I'm marth and I fsmash three people at once, does that stale it three times?

Also bump on the marth f-air question

Hitting shields does not stale moves.
Hitting Shy Guys and Brinstar's goo does.

 

[Ripple]

still waiting on my roy question from months ago

 

[ajp_anton]

Hitting Shy Guys and Brinstar's goo does.

Wait what, it does?

 

[Grim Tuesday]

If a hitbox collides with a hurtbox, the move stales. Regardless of whose hurtbox it is.

 

[Anand]

How does Stamina mode work? I've noticed attacking someone with X HP doesn't result in knockback as if they had (150-X) damage. I could play around with it myself but getting a definitive answer would be nice.

 

[ajp_anton]

Hitting Shy Guys and Brinstar's goo does.

If a hitbox collides with a hurtbox, the move stales. Regardless of whose hurtbox it is.

OK I just tested this in order to restore balance in my head.

Hitting hurtboxes that aren't part of another PLAYER (human or CPU) does NOT stale moves.
Examples of hurtboxes that don't stale moves:
- Shy(/Fly?) Guys
- Brinstar's pillars and ground
- Mushroom Kingdom's boxes
- Birdo
- Items
- Targets

 

[Kink-Link5]

I thought there were no stale moves in 1P modes.