Q&A Mechanics & Techniques Discussion

[I speak Spanish too]

can u perfect pivot out of a dash? it doesnt seem instant to me

 

[FallenHero]

can u perfect pivot out of a dash? it doesnt seem instant to me

I don't think so, but there is Dance-trotting. It allows you to do a lot of the same things that perfect pivoting can allow you to do, but it is a lot easier to learn and do.

 

[I speak Spanish too]

I don't think so, but there is Dance-trotting. It allows you to do a lot of the same things that perfect pivoting can allow you to do, but it is a lot easier to learn and do.

Not true. Dash trotting/extended dash dances/ foxtrot cancels only allows you to do anything you could normally do out of a dash.

 

[FallenHero]

Not true. Dash trotting/extended dash dances/ foxtrot cancels only allows you to do anything you could normally do out of a dash.

Yeah your right. I kind of forgot about perfect pivoting.

 

[Vipermoon]

Dancing trotting is definitely much more important for marth with his super long initial dash

 

[I speak Spanish too]

Ok, I have a question. As far as I know, you can only dash attack/grab, jump, and side-b out of the dash animation. Meaning that if you input any of the said commands while foxtrotting you can go into it in the duration of the foxtrot animation. If somebody can confirm this that would be really helpful.

Another question I have is how many frames of lag you have when you exit the foxtrot animation so you can do something else, like roll, dash again, or perfect pivot for example. Also, if you do have endlag is it significant?

 

[Azazel]

Ok, I have a question. As far as I know, you can only dash attack/grab, jump, and side-b out of the dash animation. Meaning that if you input any of the said commands while foxtrotting you can go into it in the duration of the foxtrot animation. If somebody can confirm this that would be really helpful.

Another question I have is how many frames of lag you have when you exit the foxtrot animation so you can do something else, like roll, dash again, or perfect pivot for example. Also, if you do have endlag is it significant?

you can also straight up upsmash during a dash, but then again you could also just jc upsmash as well during a dash. You can also roll during a dash but not spot dodge or shield. Since rolling instantly changes your momentum and can be done during any part of your dash (where dance trotting [extended dance dancing] is limited to the end of your dash) it's good to use it to keep your movement unpredictable. You actually see this alot in high level play, mixing extended dash dancing and rolling

for the second question it is very character dependent, with dash length and what not
the you cancel the endlag of a foxtrot with another dash which allows you to do some things earlier like F-smash, perfect pivot

 

[ぱみゅ]

I wonder if (theoretically at least) this means that intangible/invincible/teleport recoveries can hold down while recovering upwards and release the stick when they're right above the ledge to avoid any vulnerability frame.

 

[ItoI6]

does anyone know where i can get ledge get up frame data? i recall seeing a good thread on it a long time ago but i cant find it now

 

[Lavani]

I wonder if (theoretically at least) this means that intangible/invincible/teleport recoveries can hold down while recovering upwards and release the stick when they're right above the ledge to avoid any vulnerability frame.

Maybe I've done a bad job at explaining it, but that's something I've been trying to convey for some months now on the Zelda (and Sheik!) boards, and is what I show in that Sheik gif. I forgot what tournament it was but I remember seeing ZeRo do it in a set on FD some time in the last month or two.

 

[teluoborg]

Not true. Dash trotting/extended dash dances/ foxtrot cancels only allows you to do anything you could normally do out of a dash.

It's partially true actually. You can perfect pivot out of a foxtrot.

 

[I speak Spanish too]

does anyone know where i can get ledge get up frame data? i recall seeing a good thread on it a long time ago but i cant find it now

im happy you're back
http://smashboards.com/threads/lets-collect-ledge-option-frame-data.409945/

you can also straight up upsmash during a dash, but then again you could also just jc upsmash as well during a dash. You can also roll during a dash but not spot dodge or shield. Since rolling instantly changes your momentum and can be done during any part of your dash (where dance trotting [extended dance dancing] is limited to the end of your dash) it's good to use it to keep your movement unpredictable. You actually see this alot in high level play, mixing extended dash dancing and rolling

for the second question it is very character dependent, with dash length and what not
the you cancel the endlag of a foxtrot with another dash which allows you to do some things earlier like F-smash, perfect pivot

I meant your options out of extended dash dances not running. I don't know why the terminology is so similar.

 

[Dr. Tuen]

Hmm, a small pet project if people wouldn't mind helping.
A compilation of all the "best; most informative" single posts in this thread. It was apart of the original idea that cool discoveries started here and blossomed into a new thread if necessary, it's been probably saner to keep it in one place for all the scientists that visit here () but stuff has likely been buried and we can do a great service to many more if we had them more accessible.

I'm sure quite a few lavani/reflex/pikabunz/etc would be great, but yaaaaaaaaaaaaaaaaaaa~

I've actually been thinking about this too. Though less of a list of posts kind of thing and more of a referenced synthesis of information kind of thing. Example - there may be an "anatomy of a dash" section that discusses the phases of dashing and the options available out of each phase, which transitions to a discussion on the dash related techniques known to the community.

Organization, referencing, and length of the post would be interesting... But I'm up for giving it a go. I know we already have a terminology list, but I think something with more substance/explanation would benefit the community for answering basic questions. @ Shaya you're already exposed to how I format and reference posts, so you have an idea of what I'm talking about. Do you already have something started? If not, that's fine. I'm already a bit amped to go through this thread and re-read everything.

Oh, I also want to gather the various tools and equations we know about smash into that post too. Sixriver's frame data and shield advantage calculator, @KuroganeHammer 's website and work, @TourneyLocator 's option select stuff... the whole 9 yards.

 

[I speak Spanish too]

I totally agree with Tuen about the idea. My biggest concern though is anything involving frame data(shield stun, intial dash frames options, etc) I feel as though mechanics involving them need to be throughly tested and verified to avoid any type of misconceptions.

 

[Dr. Tuen]

I totally agree with Tuen about the idea. My biggest concern though is anything involving frame data(shield stun, intial dash frames options, etc) I feel as though mechanics involving them need to be throughly tested and verified to avoid any type of misconceptions.

I believe that some of those phenomena are actually well characterized, but poorly communicated. This is where a document like this would come in handy. Ideally, all of these points will be referred back to their researcher/video/post of origin so that the facts can be freely checked by those that review said document.

That said, I have the same concern as you do, but with different topics. For instance: the rage effect. Some say it starts at any percent, but to a fractional extent. Some say it starts at 50%, and so on and so forth. That would be one I'd have to check as carefully as possible.

All that said, this would primarily pertain to general mechanics and techniques. Character-specific frame data would be left to that character's frame data repositories and their parent character boards. Granted, I'll reference many of these places too, so it'll be easy to transfer a reader. For example, knowing the properties of a jump squat is definitely a part of this future document. Me knowing my main, Zero Suit Samus, has a 4 frame jump squat is up to the character-specific board. This will help reduce the size and scope of this project.

I also hope to identify problems that do not have solid solutions or explanations. This should hopefully encourage new testing or confirmation testing based on prior assertions.

 

[A_Kae]

@ Dr. Tuen

Do you want any help with that? It's a great idea, and I'd love to help out if you want.

 

[Dr. Tuen]

@ Dr. Tuen

Do you want any help with that? It's a great idea, and I'd love to help out if you want.

Thanks for the offer! Let me get some initial framing out of the way and get back to you. If anything, fact checking part way through would be really rad. For my own memory's sake, I'll jot your name down in my notes for this.

 

[Cook]

Iirc, in Brawl you could hold up on the control stick to prevent being stage spiked. For example,, if you know DK is gonna cargo throw stage spike you then you just hold up and instead of being spiked you'll bounce off the stage horizontally or diagonally. Is that still the case in this game?

 

[Vipermoon]

Iirc, in Brawl you could hold up on the control stick to prevent being stage spiked. For example,, if you know DK is gonna cargo throw stage spike you then you just hold up and instead of being spiked you'll bounce off the stage horizontally or diagonally. Is that still the case in this game?

Is that really how it was in Brawl? I thought it simply always depended on the angle of the piece of wall you hit.

 

[teluoborg]

I don't know. What I know is that in Brawl holding up when you were cargo'd by DK was the same as mashing perfectly so you could always break out of the cargo easily and escape before the stage spiking throw.

 

[san.]

I've read that the frame synching from landing with an aerial during hitlag doesn't work on shields.

Is it known why that is? Just curious.

 

[Dr. Tuen]

I've read that the frame synching from landing with an aerial during hitlag doesn't work on shields.

Is it known why that is? Just curious.

As far as I have read, the community isn't even 100% sure on that fact. It would be valuable to get better notion of frame syncing viability on shields. and if it does or does not work, try to get a reason why.

If it indeed fails to function on shields, my best hypothesis is this: landing lag frames and hitlag frames can sync, landing lag and shield stun frame cannot sync. I think it would be a system reason like that, since I don't see why a character can't land and hit shield on the same frame.

One way to test: find a move that can be punished by someone's shield grab (say, Mario's shield grab). However, pick a move that can't be shield grabbed if it's frame synced into a jab (theoretically speaking). Even better? @KuroganeHammer 's site has full hop and short hop durations. Using those, air dodges, and start up frames for certain moves, we may be able to construct a perfect frame sync. In fact... I think I saw one with Gannon somewhere... anyways, he footstooled someone, up aired, the down aired. If that is indeed frame perfect with buffering, that can be used to test this phenomenon. Potentially... if that whole, grab punish principle actually works for that move.

--------------

EDIT:

OK, I think I constructed a test.

Ganon vs Falcon, as per this video:

It looks like the inputs are footstool, up air, down air. NEW NOTE - Ganon's Dair is Electric.

If Captain Falcon has already used his shield lock frames (has held his shield for 11 frames or more), then Ganon's 19 frame hit will cause 12 frames of defender shield lag, 7 frames of shield stun, and has 28 frames of landing lag. Ganon also incurs 18 frames of attacker hitstun during this time. The defender shield lag + shield stun should subtract from Ganon's landing lag, leaving it at 15 frames disadvantage. If Captain Falcon shield drops, that's another 7 frames taken off, leaving 8 frames.

The math shakes out like this: Ganon is at a 8 frame disadvantage. Have him jab (8 frames). This means his first action after the potential frame sync is in 16 frames. Have Captain Falcon up tilt (17 frames after the shield drop is complete).

If frame sync works on shield, Ganon will always hit first. If not, Captain Falcon will up tilt just fine.

-----------

EDIT 2 - Ganon's dair is electric, had to re-do the calculations.

 

[TheReflexWonder]

One video showing things that help with general testing in Training Mode, and another on the internaction between grabs and walls--

 

[Dr. Tuen]

While I lack video evidence for my claim, due to lack of time and good internet, I have performed the test I outlined above.

1. I checked the frame sync set up in the video regarding Ganon and C. Falcon.
2. I used the 2-frame skip method described by @Fox Is Openly Deceptive to run my test. (Using two controllers was still hilariously weird)
3. I also checked the total frames between Ganon's down air hit on shield to his jab hitting C. Falcon.

Results!

• When doing Ganon's jab vs C. Falcon's up tilt as described in my test above, the up tilt won without any question.
• When counting the frames between Ganon's down air on shield and jab hits, I got 54 frames. That matches with landing lag (28) + attacker hitstun (18) + jab start up (8).
• When counting the frames of Ganon's normal frame synced down air and jab, I got 36 frames. That matches with the frame sync math: landing lag (28) + attacker hit stun (0, it overlaps) + jab start up (8).

I believe this is reasonable evidence to claim that the frame sync technique, as we currently know it, does not work on shield.

I do not, however, have a reasonable hypothesis regarding the reason for this phenomenon. Is attacker hitstun from shields different than when you strike your foe? Does something about the shield hitbox prevent you from falling to the ground and yielding the "sync" effect?

I have no idea. But, for now, I think this test can suffice until new knowledge on these sync actions (this and the no-tech phenomenon) is known.

 

[Shaya]

I think testing to see if frame syncing works on shield is by landing hit box aerials (usually dairs).
That is, if they automatically frame sync like some lead us to believe, either as on hit or on shield that would be a good base point.

 

[Ffamran]

Uh... can clanking push people off the ledge 'cause I think that just happened: https://www.youtube.com/watch?v=Lt9qUWwUHMI#t=8m41s. What I don't get is why Falco immediately enters into a Nair. Assuming Rahx only pressed up on his C-stick or pressed up and attack, he shouldn't be doing a Nair. At the very least, he should have done an Uair. Now, if he mashed attack for whatever reason, he would be buffering an attack, but even still, you shouldn't be mashing attack while doing a Smash.

 

[Jaxas]

Uh... can clanking push people off the ledge 'cause I think that just happened: https://www.youtube.com/watch?v=Lt9qUWwUHMI#t=8m41s. What I don't get is why Falco immediately enters into a Nair. Assuming Rahx only pressed up on his C-stick or pressed up and attack, he shouldn't be doing a Nair. At the very least, he should have done an Uair. Now, if he mashed attack for whatever reason, he would be buffering an attack, but even still, you shouldn't be mashing attack while doing a Smash.

Not sure, but it may be something to do with the buffering system - I know that with ZSS when ledge-cancelling Flip Kick, if you input the Kick part just before hitting the platform you'll Kick -> Land -> Slide off -> Nair with only 1 press of the A button (it also always uses Nair even if you hit a direction on the Cstick).

From what I remember, someone ( TheReflexWonder ?) was saying that platform drops into an aerial come out as a Nair (or Nspecial?) as another weird quirk of the buffer system too.

 

[Dr. Tuen]

I think testing to see if frame syncing works on shield is by landing hit box aerials (usually dairs).
That is, if they automatically frame sync like some lead us to believe, either as on hit or on shield that would be a good base point.

I'm a bit lost on this. You seem to have described what I effectively did. Unless you are talking about moves like... ZSS/Sonic dair? Diving type down airs are probably not a good choice, nor would either of the Link's dairs, since they always do the sword pull animation after. Same with DDD's up B, Yoshi's down b...

--

Somewhat related, I wonder if the attacker position can be moved during attacker hitstun. The defender gets the use of Smash DI, even if it's effect is more mild in Smash 4, so why not the attacker too? There's even shield DI, isn't there? Anyways, if that's a thing, maybe the attacker can SDI into the ground and force the sync phenomenon. If that's even remotely true, it'd extend the number of viable frames for this technique. It may even help this whole shield frame sync issue.

 

[Lavani]

I'm a bit lost on this. You seem to have described what I effectively did. Unless you are talking about moves like... ZSS/Sonic dair? Diving type down airs are probably not a good choice, nor would either of the Link's dairs, since they always do the sword pull animation after. Same with DDD's up B, Yoshi's down b...

He means moves like Falco fair or Bowser Jr. dair that have a landing hitbox. Since the landing hitboxes naturally frame sync on hit every time, it'd be very easy to test if they still function that way on shield.

 

[Dr. Tuen]

He means moves like Falco fair or Bowser Jr. dair that have a landing hitbox. Since the landing hitboxes naturally frame sync on hit every time, it'd be very easy to test if they still function that way on shield.

I don't think that's how it works. I think you must land with the aerial hitbox on the same frame as it hits to sync the landing lag and attacker hitstun. Landing hitboxes come tied to set landing animations and a set first actionable frame. It wouldn't make sense for those hits to frame sync.

Also, for an "easy" test, we need to be able to replicate the sync with frame-perfect precision. I'm not sure if I know how to do that without a setup like Ganon's. Without a set up, it's hard to say if a failure to sync on shield is because it just doesn't work on shield or if it's because I did it wrong.

 

[Lavani]

I don't think that's how it works. I think you must land with the aerial hitbox on the same frame as it hits to sync the landing lag and attacker hitstun. Landing hitboxes come tied to set landing animations and a set first actionable frame. It wouldn't make sense for those hits to frame sync.

Also, for an "easy" test, we need to be able to replicate the sync with frame-perfect precision. I'm not sure if I know how to do that without a setup like Ganon's. Without a set up, it's hard to say if a failure to sync on shield is because it just doesn't work on shield or if it's because I did it wrong.

It's exactly how it works. You'd normally have a hefty 25 frames of landing lag with Bowser Jr's dair, but on hit he's able to act almost immediately after his opponent starts being knocked back because 20(?) of those frames are eaten by hitlag. There's zero precision involved in this so it would definitely be an easy test on shield.

 

[Dr. Tuen]

It's exactly how it works. You'd normally have a hefty 25 frames of landing lag with Bowser Jr's dair, but on hit he's able to act almost immediately after his opponent starts being knocked back because 20(?) of those frames are eaten by hitlag. There's zero precision involved in this so it would definitely be an easy test on shield.

Well, color me surprised then! I guess more tests need to happen. I'll look into it something this week hopefully (if someone doesn't get there first!). I'm curious about a few things. Is it possible to actually incur those landing lag frames AND get the hit? The question we're all looking for: does it work on shield? If it's as easy to test as it seems, it'll be a good candidate for testing the attacker smash DI on shield hypothesis too.

 

[Lavani]

(if someone doesn't get there first!)

Don't mind if I do!

Here is a gfycat comparing Bowser Jr's dair on whiff, hit, and shield: http://gfycat.com/OrderlySmartIndochinesetiger

All three clips start 15 frames before landing, all three clips have shields first visible on frame 41. Frame syncing works on shield.

 

[Dr. Tuen]

Don't mind if I do!

Here is a gfycat comparing Bowser Jr's dair on whiff, hit, and shield: http://gfycat.com/OrderlySmartIndochinesetiger

All three clips start 15 frames before landing, all three clips have shields first visible on frame 41. Frame syncing works on shield.

Very cool.

However, I don't think we can say frame syncing works for all moves. The Ganon test outlined earlier failed. That could be an error, but it could also be difference between moves that have a landing hitbox and moves that do not have a landing hitbox, but I don't know.

Either way, big ups for Bowser Jr, for sure!

 

[Lavani]

I would side toward error, but my 3DS is in pretty rough shape and last time I tried looking into that with a CPU they refused to shield at the same time I'd get a proper frame synced Ganon dair for over an hour, so I'll leave it to someone with a Wii U and two controllers to test it with. I'd put money on it working on shield with standard aerials too though, I see no reason it wouldn't (and feel like I've done it on shield myself with Greninja's nair, but I have no replays to back that up with).

 

[Dr. Tuen]

I would side toward error, but my 3DS is in pretty rough shape and last time I tried looking into that with a CPU they refused to shield at the same time I'd get a proper frame synced Ganon dair for over an hour, so I'll leave it to someone with a Wii U and two controllers to test it with. I'd put money on it working on shield with standard aerials too though, I see no reason it wouldn't (and feel like I've done it on shield myself with Greninja's nair, but I have no replays to back that up with).

I used a Wii U, two controllers, and a frame perfect buffered setup in frame advance mode. I'll try to get a video, but with respect, I lean toward the idea that there are special cases. However, I'm also willing to say the the phenomenon is just not completely characterized yet.

The whole Bowser Jr case is interesting by itself. Do any other aerials behave that way? It's interesting that his frame sync happens automatically, and requires moderately less timing.

Overall, I think more thought and tests are required.

Edit - Holy crap, that 3DS has seen better days!

 

[Lavani]

The whole Bowser Jr case is interesting by itself. Do any other aerials behave that way?

Every aerial that puts out a landing hitbox should behave this way. I can confirm it for dair dair fair bair/dair dair dair at least.

The thing that might not make it very obvious is the generally low hitlag these moves tend to have,
dairs and bair are exceptions to this. dair has a 0.3 hitlag modifier, 's is just 0, and a lot of the rest are something weak like a 2% hit with no modifier so they don't generate much hitlag by default.

 

[Dr. Tuen]

Every aerial that puts out a landing hitbox should behave this way. I can confirm it for dair dair fair bair/dair dair dair at least.

The thing that might not make it very obvious is the generally low hitlag these moves tend to have,
dairs and bair are exceptions to this. dair has a 0.3 hitlag modifier, 's is just 0, and a lot of the rest are something weak like a 2% hit with no modifier so they don't generate much hitlag by default.

Interestingly, all of these examples you listed generate hitboxes on frame 1 when landing. It took a little searching, but I found one that didn't: Bowser Jr.'s forward air generates a landing hitbox that does 2% on frame 2. I wonder if that one would fail to produce the effect.

This frame 1 hit could have something to do with why this class of move works on shield. They generate a hitbox the same frame they land, but I would wager the prior utilization of the frame sync may involve hitbox generation on the frame before you land. Even the original "frame cancel" video shows Falco hitting, then landing on the next frame. But that's just me thinking out loud... I really want to go home and get back to the lab on this!

 

[Lavani]

Interestingly, all of these examples you listed generate hitboxes on frame 1 when landing. It took a little searching, but I found one that didn't: Bowser Jr.'s forward air generates a landing hitbox that does 2% on frame 2. I wonder if that one would fail to produce the effect.

This frame 1 hit could have something to do with why this class of move works on shield. They generate a hitbox the same frame they land, but I would wager the prior utilization of the frame sync may involve hitbox generation on the frame before you land. Even the original "frame cancel" video shows Falco hitting, then landing on the next frame. But that's just me thinking out loud... I really want to go home and get back to the lab on this!

I just checked it, and Bowser Jr.'s fair landing also frame syncs naturally. The basic mechanic of it is simply that landing lag overlaps with hitlag instead of being delayed by it.

I never fully understood how the game decided when to allow you to land when frame syncing traditional aerials. You don't necessarily land the frame after you hit; I saw variable timings with it, the latest I've had (garbage quality recording) was five frames after the hit, which...doesn't make a ton of sense when your character shouldn't be moving at all. Figuring that out's been something I've wondered about for months now, but I have no idea how I'd go about determining what causes that.

 

[san.]

I just checked it Bowser Jr.'s fair landing also frame syncs naturally. The basic mechanic of it is simply that landing lag overlaps with hitlag instead of being delayed by it.

I never fully understood how the game decided when to allow you to land when frame syncing traditional aerials. You don't necessarily land the frame after you hit; I saw variable timings with it, the latest I've had (garbage quality recording) was five frames after the hit, which...doesn't make a ton of sense when your character shouldn't be moving at all. Figuring that out's been something I've wondered about for months now, but I have no idea how I'd go about determining what causes that.

Nice!

So in general, aerials can be safer on shield up to the amount of self hitlag they produce on hit minus X random number?