SCOTU's Frame Data Q&A + Request Thread

[SCOTU]

the hitlag and shieldstun time depend entirely on the damage of the attack

the formulas are :

hitlag = (2 + damage/3) frames ; half of it when crouch cancelled.
shieldstun = (4.45 + damage)/2.235 frames

when you use a move a lot, its damage is reduced and so are the hitlag and shieldstun, and this makes them a bit less efficient at pressuring.

I'm guessing there're exceptions to this? but that's amazing. You should tell phanna if you already haven't.

A few requests:
With Peach's frame perfect nair pillar hitting a shield, would it be possible for an opponent to roll or jump out, or would it only be possible to grab to escape?

The timing for continuing Roy's DED depends on whether it hits or not (or so it feels like, it might just be the hitlag affecting my judgment.) If so, I think the frame data that's currently out is only for non-hitting.

And last but not least: Mr. Game & Watch's Judgment Hammer in terms of the match timer. I've always wanted to test a theory that the number shown has a corrolation to the match timer, but haven't had the frame perfect data to test it against. Would it be possible to get a few hammers done at different specific times in a match?

Thanks in advance.

Peach's "pillar" (using perfect nairs only) cannot be grabbed. It can be escaped with a roll/ spot dodge, and a jump will probably just end up in you getting hit.

Got an idea, probably been thought of before and proven impossible, however.
Dr. Mario's BTT lowest possible time is believed to be 13.40, but I'm wondering if it would be at all possible using frame perfection to get to the end of the run with enough time left as the platform is at the height (or around there) of it's path and use a frame-perfect down B to raise all the way up to meet it (or at least get on it before it's at the bottom of it's path) and then, if needed for extra height if the platform isn't at the correct height, full jump and use another frame-perfect down B to gain the rest of the height.

I don't know the methods that one would use to make a perfect BTT, I know Doraki's done several (at least) "perfect" BTT using AR, and he's probably better at coming up with the strategy for completing it.

 

[Adi]

Hey Scotu, is there any advantage for Falco using Fair or Upair in a pillar, I saw Dope doing it once and I wondered if he was just joking around or if there's less frames to be grabbed..

 

[SCOTU]

Well, while i haven't done frame analyses on fair/ uair yet, they have slower comeout times, but they have multiple hits. That means that the shield stun is likely to cover a wider range (but it might have longer lc lag). Fox's drill kick is better at shield trapping than Falco's since it has more hits, thus renewing shieldstun (although Falco's does more shield damage by a ton). Dope does tend to use the fair more than other top level falco's i've noticed.

 

[Adi]

Yeah I figured that much, just curious whether in the long run including longer LC lag and whatnot if it turned out to be superior than standard Dair pillars.

 

[Doraki]

I believe that doc btt thing has been tried by now and it isn't faster.
And about half of the AR time I got in btt eventually got beaten =/.
Stupid hitlag not decreasing, I bet.

btw, fox's drill can be shieldgrabbed by almost everyone due to ridiculously short shieldstun.
upair is better since you can have the first hit early after the shine and the second strong hit right as you land, which does a lot of shieldstun.
The same thing can be said about fox and falco's fairs I bet, but their second kick isn't as powerful as the upair's.

what's more important is your ability to begin a combo once any of your move connects.

And buffering a roll with the c-stick when your shield is pressured is ridiculously easy.
And shinegrabbing is awesome.
And So is shining out of shield.
And then, double-shining can be useful.

 

[SCOTU]

A Quick Public Announcement

It has come to my attention lately that people believe "We have found everything out". I would like to indicate to the people that believe this that they aren't being very openminded. Because we haven't figured everything out. Every so often, you find someone who has figured something new out, either about the game physcis (i.e. Phanna recently discovered you could jc a usmash by only holding up, jumping and pressing A), or some usefull aspect of a move, or some new tactic that deals with a certain style, or a new pattern that many people fall into. The game simply can't be figured out completely. Just needed to point that out. In fact, this thread is all about finding new stuff out, and analyzing found data for new insights.

 

[Falco&Victory]

Very nice thread Scotu, you've done a great job. you put in a lot of effort.

 

[Rat]

Scotu i have a quick question of hit lag and such. So during hitlag do all other animations continue? Like with the reflector if you hit them, that 3 frames of lag. Would the soonest fox can jc now be frame 8 (or 7 is the hit is calculated on the frame of the hit...) or is it still frame 4 of the shine?

Next for landing lag, is the same amount of lag applied to both jumps and wavedashes? i.e. is it always 4 for fox? I'm assuming yes...

and last, is there a list of reaction frames (or suffering from being hit frames)? Not sure of the right term for that one, couldn't find it in the term list. I mean the lag you suffer after being hit before you can do anything. ie what makes combos work. Is that what is meant by hit lag? I assumed it meant the lag after the hit before the receiver of the hit gets sent flying. If there is a list that would be nice.

 

[Jihnsius]

Does hitlag apply to all characters involved in the physical attack?

 

[Doraki]

I remember thinking about asking my teammate to come and put up his shield near me when I want to do a shine spike on the edge of a stage, so that the hitbox of the shine stays for 4 frames instead of 1.

reaction frames = stun time ?
it behaves roughly the same as the knockback of a move
the higher your damage meter and the lighter you are, the more you are stunned.

 

[Jihnsius]

it behaves roughly the same as the knockback of a move
the higher your damage meter and the lighter you are, the more you are stunned.

Stun eliminates all air resistance.

 

[SCOTU]

Scotu i have a quick question of hit lag and such. So during hitlag do all other animations continue? Like with the reflector if you hit them, that 3 frames of lag. Would the soonest fox can jc now be frame 8 (or 7 is the hit is calculated on the frame of the hit...) or is it still frame 4 of the shine?

Next for landing lag, is the same amount of lag applied to both jumps and wavedashes? i.e. is it always 4 for fox? I'm assuming yes...

and last, is there a list of reaction frames (or suffering from being hit frames)? Not sure of the right term for that one, couldn't find it in the term list. I mean the lag you suffer after being hit before you can do anything. ie what makes combos work. Is that what is meant by hit lag? I assumed it meant the lag after the hit before the receiver of the hit gets sent flying. If there is a list that would be nice.

No. During Hitlag, everything stays the same. The hitbox stays out for the shine, you can't jc it until frame 7.

Landing lag is for jumps. Each special move (and air dodges) has it's own "special landfall lag". For instance, The firefox has 3 frames of lag upon landing, and the airdodge (for everyone) has 10.

Hitlag is just both characters being "stunned" or stuck in one frame of animation, where only SDI and ASDI can take place. And yes, it's starts when the hitbox connects with the hurtbox, and ends when they go flying. It is applyied to both the attacker (or projectile) and the attackee. (although only the attackee can SDI/ ASDI).

Hitstun is what makes combos work. There is a # of frames (knockback dependant) that you cannot do anything after being hit (not including hitlag).

 

[Rat]

So to make sure I have this correct, hit stun is only dependant on the knockback of a move? Is there a complied list of how long each hit stun is?

Also with the IASA frames on moves, is it possible to turn around during them? Or turn around without the turnaround animation? Why I ask is because I was trying to connect samus' Dtilt to something behind her. Is seems that during the IASA frames, that I can't turn around. Every time I try it ends up just being a jab.

 

[SCOTU]

So to make sure I have this correct, hit stun is only dependant on the knockback of a move? Is there a complied list of how long each hit stun is?

Yep, it's only dependant on knockback (for most moves). No, there isn't a list. A while ago, i considered making one, and i found some formulas that don't really point to anything, and i realized that it doesn't really matter anyway, since you can just try stuff in training mode.

Also with the IASA frames on moves, is it possible to turn around during them? Or turn around without the turnaround animation? Why I ask is because I was trying to connect samus' Dtilt to something behind her. Is seems that during the IASA frames, that I can't turn around. Every time I try it ends up just being a jab.

I'm not 100% on IASA, but i believe you can turn around as an interrupt. You're probably just timing it too early. Samus dtilt>pivot utilt would be amazing. I'd have sex with it. But you're probably just gonna have to stick with interrupting it with a reverse fsmash/ bair.

 

[Pasqual]

Something I've been pondering:

I vaguely recall hearing that c-sticking a smash is one frame faster than analog + a. I don't really remember where I heard it but was just wondering if you knew/could check?

I'd also like to know, if there is indeed a difference, whether it also applies to c-stick charging versus analog charging.

 

[brawlpro]

i appreciate your help, thnx!

 

[SCOTU]

A Csticked smash is just an uncharaged A smash.
If you release A before a smash's "charge frame" when using A, it's the same as the cstick. The charge frame varies based on character and smash.

The charging of each is exactly the same.

 

[Rat]

Ok so another question. When using an UpB out of a shield, do you need to count the jumping frames? Or is it immediately after you press b, that the UpB comes out?

 

[SCOTU]

IT's a jump canceled Up b. So any time during the jump start animation, you can cancel the animation by pressing up-b. So the fastest up-b out of shield is
1 Start Jump (shield is down)
2 Up-B

For Samus, which i know you're asking about, the slowest up-b is
1 Start Jump (shield is down)
2
3 Up-B

And samus is invincible on the frame (and a few after) her up-b starts. So there's only 1-2 frames of vulnerability between shield and hit (of up-b). And someone like, say, Fox, will not be able to shine afterward before the up-b coming out.

 

[GamerGuitarist7]

And someone like, say, Fox, will not be able to shine afterward before the up-b coming out.

this is why i hate fighting good samus's. this also works for bowser up-b out of shield (i know you know scotu, just pointing it out for the others)

 

[Jihnsius]

scotu, can you do me a huge favor? I'm working on a Smash clone and I need to a frame count for Ness' jump data for the moment he leaves the ground 'til the moment he lands for short hop and full jump in order to fine tune how gravity and jump heights work. I'd appreciate it.

 

[SCOTU]

Yes. It's a good point. I just focused that post on samus since he mains samus, and i knew why he was asking...

But true. It works with Samus', Bowser's, G&W's - i think that's it. I'm gonna get flamed when ppl find more that i forgot...

 

[Magus420]

Peach's "pillar" (using perfect nairs only) cannot be grabbed. It can be escaped with a roll/ spot dodge, and a jump will probably just end up in you getting hit.

It actually can be grabbed by 7 framers because you never made the correction I told you about in the Peach forums . There is an extra frame between the n-air hitlag and landing lag that you are missing. It's impossible to land that fast on level ground (afaik). This lets the grab come out at the exact same time as the next n-air which gets outprioritized.

You should do ICs desynched jab cancels btw =)

 

[SCOTU]

scotu, can you do me a huge favor? I'm working on a Smash clone and I need to a frame count for Ness' jump data for the moment he leaves the ground 'til the moment he lands for short hop and full jump in order to fine tune how gravity and jump heights work. I'd appreciate it.

What exactly are you looking for, i don't really know what you're asking. If it's just timing, for how long it takes w/ & w/o FF, that's easy, else, i don't really know what you're asking. And why Ness?

It actually can be grabbed by 7 framers because you never made the correction I told you about in the Peach forums . There is an extra frame between the n-air hitlag and landing lag that you are missing. It's impossible to land that fast on level ground (afaik). This lets the grab come out at the exact same time as the next n-air which gets outprioritized.

You should do ICs desynched jab cancels btw =)

True, i did never get around to making the correction. I'll have to do that. But, (although i never tested it), i was under the impression that when a grab and an attack come out at the same time, the attack wins so long as it hits. I know grabs have a higher priority, but when they're on the same frame, and the hitbox connects, you'll be grabbed, they'll be hit, and you'll escape. I should probably check this. I'll look into IC jab cancels as well.

 

[Pasqual]

Thanks for the charging thing, scotu. I've got another question that's somewhat frame related.

I've got a friend who seems to get his shield out just a little bit faster after landing than everyone else we smash with. This has happened to the point where we've named this type of shielding after him. My question is about how he does it since, unfortunately, he doesn't know. Could there be some type of frame punishment for trying to shield too early? What I'm thinking is that you get a minor penalty of frames if you start shielding during landing lag that keeps you from optimum shield comeout speed. An equivalent of a missed tech, except closer to 4 or 5 frames instead of 20 before the action becomes viable again.

Edit @ Jihnsius: If I'm reading correctly, I think you're asking for the frame counts that can be found in SDM's frame data.

 

[SCOTU]

@ Pasqual: I'm tempted to say he's L-canceling and you're not. But that would be an unfair judgement of your skill. No, there's no special technique. The R/L button's "buffer" as in, If you hold one of them down, the shield will come out as early as possible. So if you shffl a nair (for example), l cancel it, and during the laning you hold R, your shield will come out just as fast as someone who perfectly times it after landing from the shffld nair. except that they'd have to time it perfectly, where you just hold R.

@Jihnsius: You should invest in an AR if you're going to make that smash clone, and moreso than simulating physics with equations of some sort, you should just have lists of pre caculated heights at ceratain frames for jumps. There are equations for being hit, however, but for the most ssbm-like feel, it'd be better to just do everything as a non-continuous frame by frame representation (as your simulation could then be perfect). I also left some other notes in your thread that you may want to consider.

 

[Pasqual]

Nah, it's not an l-cancel thing. Thought about it but this is in comparison to the shield comeouts of multiple players both better and worse than us in most aspects of the game, and also after air dodges/upB/whathaveyou. I understand the buffering but I was just thinking there mighta been a couple frames of lag on it when compared to shielding the first frame out of landing lag. I guess I'll just chock it up to something that happened a few times and got blown out of proportion or something. Thanks anyway.

 

[Jihnsius]

What exactly are you looking for, i don't really know what you're asking. If it's just timing, for how long it takes w/ & w/o FF, that's easy, else, i don't really know what you're asking. And why Ness?.

Yeah, just timing for jumps with and without fastfalling. I'm using Ness as a basis for this project because he's an all-around character in terms of stats. And not to mention his spriteset is easy to come across.

 

[Zankoku]

Actually, Ness has some of the strangest jump physics in the game. You're better off starting with Mario or Doc. I think the guy doing his knockback calculations used Doc as a base because his physics were the same in NTSC and PAL.

 

[PDOT]

How fast does Ylinks Up+B hitbox come out?

 

[Magus420]

But, (although i never tested it), i was under the impression that when a grab and an attack come out at the same time, the attack wins so long as it hits. I know grabs have a higher priority, but when they're on the same frame, and the hitbox connects, you'll be grabbed, they'll be hit, and you'll escape. I should probably check this. I'll look into IC jab cancels as well.

The attack hitbox never actually hits because a grab that connects is handled first (hence the priority), so they are no longer attacking anymore at that point. If they both connect on the same frame the attacker gets grabbed and the grabber does not get hit at all since there's nothing to hit them with.

 

[SCOTU]

Yeah, just timing for jumps with and without fastfalling. I'm using Ness as a basis for this project because he's an all-around character in terms of stats. And not to mention his spriteset is easy to come across.

Jump: airborne on frame 5

Air time: 58
Earliest FF: 30
FF air time: 45

SH air time: 34
Earliest FF: 19
SH FF air time: 24

(See next comment)

Actually, Ness has some of the strangest jump physics in the game. You're better off starting with Mario or Doc. I think the guy doing his knockback calculations used Doc as a base because his physics were the same in NTSC and PAL.

He speaks the truth. Ness does have one of the weirdest jumping mechanics in the game, and Dr Mario would be a much better choice, gameplay wise, but the spriteset is far more available for ness, i'll give you that.

How fast does Ylinks Up+B hitbox come out?

They both hit on frame 8 on both the ground, and in the air.

The attack hitbox never actually hits because a grab that connects is handled first (hence the priority), so they are no longer attacking anymore at that point. If they both connect on the same frame the attacker gets grabbed and the grabber does not get hit at all since there's nothing to hit them with.

Interesting, I'll have to keep that in mind.

 

[Emblem Lord]

Priority rules in Smash are stupid.

I hate them.

 

[SCOTU]

I knew the priority rules, i just thought they applied only to overlapping hitboxen.

 

[Emblem Lord]

I have a request for Marth.

Marth can Auto cancel his nair. Just short hop, then nair, then fast fall as soon as the second slash starts. The lag is 4 frames. I want to know what options Marth has to stuff an opponent's retaliation after an auto cancelled nair.

The AC Nair wold hit with the sweetspot so it has the most stun. Can Marth do anything after this to stuff a shield grab?

 

[Jihnsius]

Thanks, scotu.

 

[SCOTU]

I have a request for Marth.

Marth can Auto cancel his nair. Just short hop, then nair, then fast fall as soon as the second slash starts. The lag is 4 frames. I want to know what options Marth has to stuff an opponent's retaliation after an auto cancelled nair.

The AC Nair wold hit with the sweetspot so it has the most stun. Can Marth do anything after this to stuff a shield grab?

Assuming the earliest possible hit of the nair (as if the first hit hit), you cannot avoid it.
Assuming the latest possible hit of the nair, all you can do is jab.

 

[SCOTU]

Public Service Announcement

Pillaring is a no good tactic that is not safe, and is a winmore: if your pillaring is helping you, they're not good enough for you to need to pillar to win. if you shffl into their shield, it's best to shine-grab, or shine>retreating rshl.

 

[Adi]

Hey Scotu if you responded or not, but could you possibly find out the time a shield is not in stun if you're doing a falco pillar except substituting a U-air or F-air for the down air.

 

[SCOTU]

yeah, i've got a bit to figure out still, but some of it's probably gonna have to wait for friday or even monday, since i'm getting ready to move back to school this week - explaining some of the lack of new frame data. Sorry for the inconvenience.