Q&A Mechanics & Techniques Discussion

[Vipermoon]

Clamhat possibly discovered the other day that gravity doesn't affect hitstun. This is interesting to me, discuss.

Lavani you might be interested in this.

Just to make sure, you're saying that when knockback gets adjusts to account for gravity, the hitstun doesn't change with it?

If so, this doesn't sound like a big deal because one can air dodge on F41 anyway.

 

[Lavani]

Just to make sure, you're saying that when knockback gets adjusts to account for gravity, the hitstun doesn't change with it?

If so, this doesn't sound like a big deal because one can air dodge on F41 anyway.

That's not happening until you're taking 100 units of knockback though. It would've been relevant for the 80~99KB range.

 

[ぱみゅ]

So I've been asked a lot about the Z-axis.
Is there any way to know what moves get affected by it?
Or is just experimenting?

 

[Lavani]

So I've been asked a lot about the Z-axis.
Is there any way to know what moves get affected by it?
Or is just experimenting?

Mainly experimentation and observation. If you can visibly see a character('s limb) moving into the foreground/background that's a pretty good clue on its own. If you see a move whiff when it looks like it should've hit, replicating the scenario on a 2D stage (Duck Hunt, Flat Zone Ω, etc) or pausing and panning the camera for a side shot can fill you in more as to whether it's a Z-axis issue (Greninja fsmash/dsmash/fair, DK ftilt) or just a naturally existing blind spot (Zelda ftilt).

 

[KuroganeHammer]

 

[Vipermoon]

So Lavani , I know we established that Limit Cloud's higher gravity gets factored into his knockback adjustment normally.

But... what happens when Cloud doesn't have Limit, gets hit or thrown, and taking that damage gives him Limit? In that instance before launch, does his fall speed and gravity "cheat" the system or is it still accounted for?

 

[KuroganeHammer]

>Cloud gets hit
>KB is pre-calculated
>Cloud gets limit

Cloud dies to Rosa utilt at the same percent without limit as he does on the hit that would give him limit.

 

[Radical Larry]

Spoiler: Very Compact Post about Pivot Forward Smash Mind-Game and Tech

The Pivot Forward Smash Attack; it's a very hefty and useful technique isn't it? Well, it's a very hard one at that, and should be used in professional levels of the competitive scene. I'm going to elaborate on how to perform it, which characters benefit most and if it's tournament viable or not.

Now that is a Pivot Forward Smash Attack? Well, it's just as it sounds, a Forward Smash that is pivoted, but like a Pivot F-Tilt and Pivot Grab. The key difference, however, is that this is actually much tougher to perform than either, but it has more beneficial results than a Pivot F-Tilt, depending on who you're using of course. To perform the Pivot Forward Smash, or PFS, all you have to do is keep a dash going (not an initial dash) and once you turn, around 1-2 frames after it, you input a Smash Attack in the opposite direction.

Using this technique actually keeps the opponent guessing! This is a type of mind-game that opponents won't know how to deal with and at some points, can be more beneficial than using a Pivot Grab depending on the character.

Spoiler: Characters who can most Benefit

Now every single character can perform this, but there are many characters who can benefit the most from it. The characters that benefit the most from it are the following types, and they can be combined: Long Reach Smashes, Quick F-Smash, Low Traction and High Speed. So which characters are the best at using this type of technique? Well, here's the top 10 characters who can perform it, in order of the benefits:

Marth is every single one of these combined; he has low traction, a quick F-Smash, a long reaching F-Smash and a pretty decent speed. Using the Pivot F-Smash is like using a Wavedash > F-Smash from Melee, it has that sort of feel. It can be used during a retreat or going behind the opponent.

Lucina is again, everything combined, but since she doesn't have the tipper, she has less benefits of this than Marth. It can be used during a retreat or going behind the opponent.

Mewtwo's attributes are everything combined, and are even like Marth and Lucina, albeit with a bit less range. His low traction and extremely high speed allow him to space opponents quite well. It can be used during a retreat or going behind an opponent.

While Luigi is slower than the two aforementioned characters, but he has a quick F-Smash and very low traction, allowing him to easily utilize this to the biggest benefit; also, his F-Smash is pretty powerful. Due to his low mobility, it's best to use this during a retreat.

Falco is slow as well, but he surprisingly has somewhat low traction and a longer-reaching F-Smash. It's best, however, to make sure that Falco is always retreating from the opponent.

Shulk can use it normally and with Speed Monado, and thanks to his long reach with his F-Smash and his somewhat low traction and decent mobility, he's able to use this effectively. What makes this far better is when Speed or Smash Monado are used, where Shulk's movement allows him to move behind an opponent instead of needing to retreat, or he can use it to quickly kill the opponent.

/ The Pits have pretty low traction, a fast F-Smash, a long reach and great mobility; what's better than that? This is best used with retreating than going behind the opponent, however, since the Pits don't have that fast of a mobility.

Link doesn't have the best of mobility, but his F-Smash comes out somewhat fast and has a very long reach. Depending on the distance of Link and the opponent, Link can use his first hit or his second hit, depending on if he can tipper the first hit. This can only be used with a retreat, as Link would never be able to get through opponents with his low mobility.

A slow forward smash, but very long reach accompanied by a fast speed, he definitely benefits from using this technique and mind-game. Since he can have his speed on him, he can retreat or go past the opponent to use this.

Now despite all of this, there are some some characters I highly recommend you do not use this with, and they are:


This is either due to very low speed altogether, how their attacks are designed, the speed of the attacks overall, their traction or the fact that one character stops their own traction with the F-Smash.

I really do find benefit from using this in ground play since it's an amazing mind-game tactic to make your opponents now guess what's coming next. But hey, I'd like to know any opinions on this; would any of you use this in competitive play given the usage of training and time?

And do any of you know about this beforehand?

 

[Lavani]

either inconsequential or a super old change

following up on this to confirm deku nut hitstun is normal in 1.0.4

mucking around with other attacks, it seems like hitstun was nerfed in 1.0.4. Hitstun on things in 1.0.0 seems roughly 20~25% higher, though it only affects jump (airdodge/attacking out of hitstun still possible as they currently are)

fun fact, I guess

 

[Vipermoon]

following up on this to confirm deku nut hitstun is normal in 1.0.4

mucking around with other attacks, it seems like hitstun was nerfed in 1.0.4. Hitstun on things in 1.0.0 seems roughly 20~25% higher, though it only affects jump (airdodge/attacking out of hitstun still possible as they currently are)

fun fact, I guess

Nice find. What about specials out of hitstun?

 

[KuroganeHammer]

oml 1.0.0 is literally a different game

I think I'll start running 1.0.0 side tournaments

 

[Kofu]

oml 1.0.0 is literally a different game

I think I'll start running 1.0.0 side tournaments

Greninja Hydro Pump for top tier

 

[KuroganeHammer]

I've collected some data on using actions during hitstun. I tested this by using Pikachu's straight ftilt on another Pikachu at various percents. The number is the first frame you can use that action during knockback.

| 50% | 60% | 70% | 80% | 90% | 100% | 110% | 120% | 130% | 140% | 150%
Jump + Specials | 34 | 37 | 41 | 44 | 48 | 51 | 55 | 58 | 62 | 65 | 68
Aerials | 34 | 37 | 41 | 44 | 46 | 46 | 46 | 48 | 53 | 58 | 63
Air Dodge | 34 | 37 | 41 | 41 | 41 | 41 | 41 | 41 | 44 | 49 | 54

This is very old data but now that we have the KB calculator we should probably begin to re-examine it.

Does anyone have any ideas on testing this? What interests me is the threshold stop, then the threshold increase.

 

[Yikarur]

There is no 1.0.0 on the WiiU thoug. I'd like to play the very first 3DS Version on the WiiU so badly.

 

[A10theHero]

This is very old data but now that we have the KB calculator we should probably begin to re-examine it.

Does anyone have any ideas on testing this? What interests me is the threshold stop, then the threshold increase.

This is what I've been doing to find the frame leniency on Up Throw to Turnaround Thunder, though I don't know if it'll be useful or not:

Spoiler

Basically, what I do is use Pikachu's Up Throw on a character at various percents and have them air dodge as soon as possible with my Turbo Controller. I then save the replay, upload it, and count the frames it takes for the air dodge to come out. (On YouTube you can advance one frame at a time using the ">" key)

An example:

For consistency, I start counting from the first frame the opponent leaves Pikachu's head.

And then I count frames until the air dodge animation starts.

So at 115%, there are 22 * 2 (44) frames from the moment she’s off of Pikachu’s head to when her air dodge animation starts (not when the F5 intangibility starts). Bat Within gives Bayonetta a F1 air dodge anyways, so leaving it at that is fine in this case. (And then it was determined separately that from that same first moment, Pikachu needs about 22 frames before he can start running. Thus there are 22 frames to input the combo.)

An obvious limitation of this method is that replays are uploaded at 30 fps so there are a few frames of uncertainty (this can be accounted for if someone has a capture card that records at 60 fps). A more minor issue is that freshness has to be taken into account since replays can't be created in Training Mode.

It's important to note that the place I chose as a starting point wasn't necessarily when hitstun starts. To figure that out, we could use Pikabunz's data and set the situation as Pikachu using a straight ftilt on another Pikachu. F41 is when he could first air dodge, so we could just count 41 frames backwards to see when hitstun begins.

Once the starting point has been determined, you can use the calculator to find out how many frames of hitstun there actually is and compare that to the frame when air dodging is possible under the same circumstances.

I hope this helps!

 

[Clamhat]

This is very old data but now that we have the KB calculator we should probably begin to re-examine it.

Does anyone have any ideas on testing this? What interests me is the threshold stop, then the threshold increase.

My first instinct is that there has to be some value in battle_objects that's affecting this. Just looking at that table, there are some things I notice:
- Jumping/Specials appear to have no threshold whatsoever
- once the threshold for airdodge/aerials is crossed, it appears to increase by 5 frames for every additional 10%, even though the hitstun from the calculator only increases by about ~3-4 frames (if both continued increasing at the same rate, air dodge hitstun would catch up to actual hitstun at around 230%

I have a turbo controller at home so I'll definitely be testing this as soon as I can, but I feel the need to point out that if a bunch of different people start labbing the same thing, we're going to need to standardize how we talk about this stuff as much as possible. Stuff like how we describe FAF, hitstun, hitlag, which frame we consider events to occur on, etc. Miscommunication will just give everyone more work to do.

 

[Clamhat]

In the next few days (probably after my finals are over) I plan on making a "Knockback/Hitstun/Etc." megathread to explain what we know/what we don't know and to centralize discussion of the topic. If you want to talk to me about it in the mean time, you can tweet/DM me or find me on Discord (Clamhat#9504).

 

[Lavani]

Fun fact of the day: knockback caps at 400 units, with hitstun capping at 160f as a result.

 

[Vipermoon]

If freshness is a 1.05 damage bonus then why does Marth's Critical Hit do 64% instead of 63% in both Brawl and S4? Base damage is 60 (tribute to the max 60 hp in his games lol). It would have to be 1.067 fro 64%.

 

[Lavani]

If freshness is a 1.05 damage bonus then why does Marth's Critical Hit do 64% instead of 63% in both Brawl and S4? Base damage is 60 (tribute to the max 60 hp in his games lol). It would have to be 1.067 fro 64%.

It doesn't.

At least not on its own. If there's other decimal damage in place from other attacks, then 64% is possible.

 

[Vipermoon]

It doesn't.

At least not on its own. If there's other decimal damage in place from other attacks, then 64% is possible.

I accounted for decimals in my question. What I didn't do was test it; I was going off memory. But man during my casual Brawl days many years I thought it did 64 on fresh stocks. Just tested it in Brawl and Smash 4 and I was wrong. The random question came up because I was wondering which moves would even need to be capped at 400 knockback...

 

[LancerStaff]

I was wondering which moves would even need to be capped at 400 knockback...

Less "which moves" and more "which modes." Probably something like a heavy giant metal smash with ultra powerful and defensively crippling critical hitter equipment on a Buster Shulk against another one with smash using power vision and max rage. Not that I've tried this or anything...

 

[prestonpreston]

By playing with my friends we got to replicate that by theching very close to the ledge, you can just have your stick tilting to a side and your character will bounce left or right to the ledge and just grab it. Also if doing it on the platform it works as a tech-cancel, which you tech but continued in the air next to the platform just as if you ran out of it. I dont know why it happens instead of a roll or just teching on the ground normally. Also, i found that if you do an aerial after running of the ledge instatly, you will get a nair (i havent tested it with other aerials actually). Just because you are running you have your stick to the left or right for a fair. But if done just at the time you leave the ledge, you will to a nair, works with platforms too.

 

[Xermo]

These aren't anything new or exclusive to smash 4.

 

[prestonpreston]

These aren't anything new or exclusive to smash 4.

ok thanks XD

 

[Bmacster]

Just a simple question, is it harder to wall tech in this game than it is in melee and if it is why?

 

[Konneh]

Now that is a Pivot Forward Smash Attack? Well, it's just as it sounds, a Forward Smash that is pivoted, but like a Pivot F-Tilt and Pivot Grab. The key difference, however, is that this is actually much tougher to perform than either, but it has more beneficial results than a Pivot F-Tilt, depending on who you're using of course. To perform the Pivot Forward Smash, or PFS, all you have to do is keep a dash going (not an initial dash) and once you turn, around 1-2 frames after it, you input a Smash Attack in the opposite direction.

I know I'm super late to reply to this post, but if anyone has trouble using a pivot FSmash, I suggest you try the "A+B" Smash Input. Just activate the A+B Smash Input in your control scheme and try doing it with both buttons. A ton of people at my locals who had trouble with the technique learned it extremely quickly and told me they find it a lot easier inputting it that way.

 

[Lavani]

Been doing a lot of research on autolink angles in the past 24 hours. Some interesting stuff of note:

• I jumped the gun on the "knockback caps at 400 KB" thing. My bad. Hitstun does appear to cap at 160f regardless, though. No nevermind that's actually also false, I have no idea how I managed to get 160f in triplicate testing on a move with 255 KBG before. MY BAD.
• The nerf to autolink angles in 1.0.4 didn't weaken downward knockback, it terminates knockback immediately after the victim drops a set distance. This does not terminate hitstun, however.
• Autolinks have an identical limit to upward knockback.
  • Likely this is to prevent ridiculous combos, as autolinks could otherwise provide a source of high-hitstun launching that keeps the opponent in range. It could also be to keep opponents from flying out the top of autolink moves.
  • This cap existing also explains why so many moves with autolink angles use a high power vertical first hit instead of autolinks all the way through ( upBs, etc)
• 367° is the only angle that does not have a cap on downward knockback. The upward cap is present, however.
  • Seeing as 367° is only present on windboxes and certain customs that wouldn't really be abusable in such a manner (Shocking Skull Bash, low-KB multihits of Chain Judge), lethal autolink spiking like in 1.0.0 isn't possible regardless.
• The knockback height limit can be circumvented if the opponent bounces off a wall.
• 365°/367° hits that don't lift the opponent off the ground can result in stationary grounded hitstun that still scales to the would-be knockback of the move. In a hacked environment this can result in effects similar to Ryu lv3 Focus Attack without the knockdown. Unsure (but doubtful) of any real-game applications.
• 366° causes vertical knockback against a grounded opponent instead. While it was originally believed this was just set to 85°, it seems to scale down to 45° in reverse Sakurai angle fashion. More research needed.

 

[Vipermoon]

Ryu's first hit of Uair is apparently 367°, perfect for Shoryuken...

 

[Lavani]

Ryu's first hit of Uair is apparently 367°, perfect for Shoryuken...

I was thinking I probably missed one by glossing over DLC characters.

It's kept in check by 20 WKB though, vs Sheik it has the same amount of hitstun as it does landing lag and that's never going to scale up (outside of rage, but 14 x 1.15 = lol)

 

[Sonicninja115]

Can someone explain to me the mechanics of Yoshi's DJ? I have been thinking on it recently, and I have a few thoughts. So, does the super armor come out frame 1? Why can he escape Mewtwo's Fthrow and does anyone else have Super armor that would allow them to do the same? Cloud? If Yoshi can get out, couldn't Bayonetta BW out?

 

[Zapp Branniglenn]

Can someone explain to me the mechanics of Yoshi's DJ? I have been thinking on it recently, and I have a few thoughts. So, does the super armor come out frame 1? Why can he escape Mewtwo's Fthrow and does anyone else have Super armor that would allow them to do the same? Cloud? If Yoshi can get out, couldn't Bayonetta BW out?

Yes Yoshi's double jump armor is active on frame 1. As for how long it lasts, I remember somebody proposing that it lasts for frames 1-69 (70 frames total). The reason Yoshi can act at all to perform double jump is because of his weight. Mewtwo's Fthrow is weight based, so it's animation takes longer to perform on a heavyweight characters. The shadow balls appear later, giving Yoshi the one frame he needs to escape with double jump. Try Fthrowing different characters in training mode. On lighter opponents like Bayonetta and Cloud, all six hits register. On Yoshi. the five shadow balls register a combo, but not the initial throw which would make it six hits. And on characters as heavy or heavier than Yoshi, they also have frames to escape. But with no frame 1 super armor or invulnerability move, it doesn't matter. Bowser probably has as many as three frames to act, but his garbo air dodge is active on frame 4, so he can't escape.

It's possible that more heavyweights can escape by using their air dodges at specific, later % ranges, but you'd need to do extensive testing to find out. And the knockback for this throw doesn't seem to scale much at all.

Come to think of it, if Ryu were just a hair heavier, he would maybe be able to escape with focus attack.

 

[KuroganeHammer]

something something http://kuroganehammer.com/Smash4/Yoshi

 

[EquenimeusW]

Hey, so i may have found something but i'm not sure about it. So on the non-omega jungle japes stage, with all tether recovery characters except toon link and lucas, if one goes to the rightmost ledge and tethers from a certain distance, then jumps, their jump will go much higher in the air than a typical ledge jump, but if preformed just right, it makes the character slide across the central platform at a high speed while being able to act. im not sure what kind of properties the river has but it seems like a high power windbox, but im not sure. I just thought it'd be worth mentioning. i found it while messing around in training and here's an example of it (sorry for low quality recording) https://www.youtube.com/watch?v=yqWdGYpdst8
idk what uses it would have being non-omega and all but i just thought i'd share

 

[Sonicninja115]

something something http://kuroganehammer.com/Smash4/Yoshi

I looked it up first, but didn't understnad the DJ thing. I wanted to be sure before making a video and possibly getting hazed.

Yes Yoshi's double jump armor is active on frame 1. As for how long it lasts, I remember somebody proposing that it lasts for frames 1-69 (70 frames total). The reason Yoshi can act at all to perform double jump is because of his weight. Mewtwo's Fthrow is weight based, so it's animation takes longer to perform on a heavyweight characters. The shadow balls appear later, giving Yoshi the one frame he needs to escape with double jump. Try Fthrowing different characters in training mode. On lighter opponents like Bayonetta and Cloud, all six hits register. On Yoshi. the five shadow balls register a combo, but not the initial throw which would make it six hits. And on characters as heavy or heavier than Yoshi, they also have frames to escape. But with no frame 1 super armor or invulnerability move, it doesn't matter. Bowser probably has as many as three frames to act, but his garbo air dodge is active on frame 4, so he can't escape.

It's possible that more heavyweights can escape by using their air dodges at specific, later % ranges, but you'd need to do extensive testing to find out. And the knockback for this throw doesn't seem to scale much at all.

Come to think of it, if Ryu were just a hair heavier, he would maybe be able to escape with focus attack.

Thanks!

 

[Pixel_]

Also Yoshi's DJ is heavy armor, not super armor; it has a limit for what it can armor through.

 

[KuroganeHammer]

I looked it up first, but didn't understnad the DJ thing. I wanted to be sure before making a video and possibly getting hazed.

Thanks!

U could have just asked me directly if you wanted to

 

[Goombo]

Since I was experimenting with footstools a bit, I've come up with some questions:

Do different charakters have different "hurtboxes" (or mor accurate: hurtbox-sizes) for footstools?
Do visualisations of those exist?
And what is the Number of frames until you can act out of getting footstooled? Is dependant on something or always one constant framevalue?

 

[Pixel_]

How much does Bowser Jr's crouch cancel reduce damage and knockback? Because of the clown cart, that makes his crouch cancel the best in the game, right?

Oh yeah and does crouch cancelling work while crawling?

 

[Masonomace]

How much does Bowser Jr's crouch cancel reduce damage and knockback? Because of the clown cart, that makes his crouch cancel the best in the game, right?

Oh yeah and does crouch cancelling work while crawling?

I don't know how much clown kart + crouch canceling reduces, but I know that Monado Shield art for Shulk makes his crouch cancel the absolute best in normal vanilla meta.

0.67x = damage taken multiplier
0.78x = knockback received multiplier
0.85x = crouching knockback multiplier(?)