Smash Ultimate Frame Data (E3 Demo)

J0eyboi

Smash Ace
Joined
Dec 28, 2017
Messages
534
#41
I think I’m speaking for all of us when I say that you shouldn’t sweat it so much... You’re doing great work man.

The Fsmash buff... The Sakurai patented “let’s buff this character but not really lol” buff. Probably doesn’t mean anything on shield because of the (lack of) shieldstun, it should still be punishable by nearly everything. Barely matters on whiff because that’s still a solid half a second+ endlag. Dunno. Kinda hope they buff something really stupid in there like Dsmash with the rest of it lol.

Hey, have other Fsmashes been buffed? If others (perhaps Link’s, that’s also two hits) that could be evidence that the shieldstun changes are intended and are being designed around.
Nope. Cloud's got nerfed for some reason, it has 3 frames more endlag (it already had 35 frames of endlag so I'm not sure why). In Pit's case, it's probably more a QoL change than anything, to make the move feel better.
 
Last edited:

Thinkaman

Moderator
Moderator
Joined
Aug 26, 2007
Messages
5,904
Location
Madison, WI
3DS FC
1504-5749-3616
NNID
Thinkaman
#42
Nope. Cloud's got nerfed for some reason, it has 3 frames more endlag (it already had 35 frames of endlag so I'm not sure why).
Cloud nerfs are about as predictable as it gets; this nerf makes sense as Cloud is disproportionately a problem at lower level play. (While still being a top ~3 character at top-level)
 
Joined
Apr 13, 2014
Messages
1,636
Location
Santa Ana, CA
#43
Hey, have other Fsmashes been buffed? If others (perhaps Link’s, that’s also two hits) that could be evidence that the shieldstun changes are intended and are being designed around.
Nothing I can confirm for F-smashes specifically. Kirby's seems to go a bit farther, which should help it cross up. Range in general and killing opponents outright are two of Kirby's core weaknesses so I could see it getting more use along with Usmash that comes out two frames faster. All of Pac Man's smashes got a startup reduction of 2-3. Sheik's second hit of Fsmash comes out 6 frames faster, and the total duration is 6 frames faster to correlate. Cloud's appears to have an endlag increase of 3 frames as far as I've seen. Otherwise fsmashes have looked very normal. I hope smashes become more relevant, but now that you can do them in either direction from a run they're probably very iffy on tweaking ground options to the same level that they've reduced aerial landing lag. Personally I'd take a wait and see approach as well.

Having grounded moves do more shieldstun is a cool idea for balance in a smash 4 meta sense. It's exceptionally rare to block a grounded attack however. Most grounded moves come out so fast that a human cannot possibly react due to input lag and general sensory perception. Only the longer startup smash attacks, charged smash attacks, or projectiles can be expected to be blocked consistently by an attentive human player. Aerials, on the other hand, have more noticeable tells. You have to be airborne to use them, so an opponent jumping toward you is either coming down with an aerial or doing an empty hop fakeout. Even the latter is rare since many smash 4 characters had at least one long range, low landing lag, high damage aerial that was safe to land with, so that's what their opponent would be anticipating with their block and punish. When players block a rogue tilt, they're generally not as expectant that they will block something and already working on their punish during shieldstun like they would be against an aerial. Hence why getting good examples of shieldstun in ultimate through footage of matches has been so difficult. If I had the game in my hands, a hand on each controller, a shieldstun test would be easy since I'm not guessing whether my shield will block an attack, I made the attack happen.

Also lol at the possibility of a demo being announced on wednesday's direct. I'd assume it would be the E3 build with only online versus matches being allowed. So a lot of unfortunate souls are gonna just see me performing moves in my own corner to get data and not really fighting. Can even upload my footage of this as sources for my charts. Fingers crossed.
 
Last edited:

J0eyboi

Smash Ace
Joined
Dec 28, 2017
Messages
534
#44
Saw this post on Reddit today. Most of it's not really relevant to here, but the OP brought up something interesting about shieldstun.

According to him, he couldn't drop shield during a multihit jab, which is interesting, because it goes against my assumption that the demo was using Brawl shieldstun. Seems like shield stun is either more complex than in previous games, or broken.

Also apparently the guy tested the roll staling which might be useful.
 
Joined
Apr 13, 2014
Messages
1,636
Location
Santa Ana, CA
#45
Saw this post on Reddit today. Most of it's not really relevant to here, but the OP brought up something interesting about shieldstun.

According to him, he couldn't drop shield during a multihit jab, which is interesting, because it goes against my assumption that the demo was using Brawl shieldstun. Seems like shield stun is either more complex than in previous games, or broken.

Also apparently the guy tested the roll staling which might be useful.
They neglected to mention which jab they tested. But in general all or nearly all of the multijabs have been changed in Ultimate to strike faster at lower damage. Usually two frames between hits which would indeed make it a true block string since the minimum shieldstun per hit was always 2 in Smash 4. This should also result in multijabs being harder to escape with just basic DI, yet easier to escape with SDI if that ever gets implemented. A lot of multihit moves have seen similar adjustments to hit timing as well. But we have confirmed that there is no restriction on perfect shielding the final hit of a multihit move. I've got my fingers crossed that this is an oversight they will fix in the final build, since I don't want multihit attacks that aren't perfect block strings to be total liabilities in competitive play due to their weakness to being perfect shielded with practicable timings.
 

J0eyboi

Smash Ace
Joined
Dec 28, 2017
Messages
534
#46
They neglected to mention which jab they tested. But in general all or nearly all of the multijabs have been changed in Ultimate to strike faster at lower damage. Usually two frames between hits which would indeed make it a true block string since the minimum shieldstun per hit was always 2 in Smash 4. This should also result in multijabs being harder to escape with just basic DI, yet easier to escape with SDI if that ever gets implemented. A lot of multihit moves have seen similar adjustments to hit timing as well. But we have confirmed that there is no restriction on perfect shielding the final hit of a multihit move. I've got my fingers crossed that this is an oversight they will fix in the final build, since I don't want multihit attacks that aren't perfect block strings to be total liabilities in competitive play due to their weakness to being perfect shielded with practicable timings.
2 frames between hits actually wouldn't be a true blockstring, you can drop shield the same frame the next hit comes out. It would have to be every other frame to be true. A lot of them were reduced to every other frame, but that's not what you said and I'm pedantic. Also the game using Smash 4 shieldstun doesn't line up with the other numbers, and we've seen at least one move (Ike's Nair in that one clip) cause less than 2 frames of shieldstun.
 
Last edited:
Joined
Apr 13, 2014
Messages
1,636
Location
Santa Ana, CA
#47
2 frames between hits actually wouldn't be a true blockstring, you can drop shield the same frame the next hit comes out. It would have to be every other frame to be true. A lot of them were reduced to every other frame, but that's not what you said and I'm pedantic. Also the game using Smash 4 shieldstun doesn't line up with the other numbers, and we've seen at least one move (Ike's Nair in that one clip) cause less than 2 frames of shieldstun.
Right, "hits every two frames" is how I should have worded it, my mistake. While I'm convinced shieldstun is overall less than in Smash 4, that Ike N-air example I found earlier supports my other theory that shieldstun simply doesn't work in that demo build and you can jump out of it. We also saw a jump out of the shield drop frames as well. What a mess. I'm willing to believe Smash 4's two frame minimum rule will still be in effect until I see more concrete and varied evidence that doesn't involve that buggy shield jump.
 
Joined
Jan 28, 2014
Messages
7,809
Location
Buried under 990+ weapons
3DS FC
1504-5709-4054
#48
I’m still not convinced that jumping out of shieldstun is a bug. Wasn’t it something we saw several Treehouse members use? It makes me think it was something present for a long time, at least. Unless there’s been other revelations but I’m assuming no.

Being able to perfect shield multihit moves seems like the exact kind of oversight/questionable game design that we’ll have to live with for a while... Potentially until next Smash.
 

J0eyboi

Smash Ace
Joined
Dec 28, 2017
Messages
534
#49
I’m still not convinced that jumping out of shieldstun is a bug. Wasn’t it something we saw several Treehouse members use? It makes me think it was something present for a long time, at least. Unless there’s been other revelations but I’m assuming no.
First of all, none of the Treehouse members used it, to my knowledge. Second, no one was using it intentionally or consistently. Third, being able to jump out of shieldstun would likely make shielding a stronger option than it's ever been before, especially with the reduced pushback, and somehow I doubt that was their goal.
 
Joined
Jan 28, 2014
Messages
7,809
Location
Buried under 990+ weapons
3DS FC
1504-5709-4054
#50
First of all, none of the Treehouse members used it, to my knowledge. Second, no one was using it intentionally or consistently. Third, being able to jump out of shieldstun would likely make shielding a stronger option than it's ever been before, especially with the reduced pushback, and somehow I doubt that was their goal.
The Zelda up against Ike was Treehouse staff, right? They weren’t using it consistently, no, but Treehouse employees have proven themselves to not be the most coordinated players out there. In a game where pressure would be higher than ever before, making shields that much better would make sense coming from the mind that thought Brawl hitstun canceling was a good idea.

I have a hard time believing it’s a bug simply because A: It exists here in the game and B: I don’t see how or why it would be introduced because shield mechanics aren’t that much different than before. It’s something that pairs well (makes it even more problematic?) with the lower and standardized jumpsquats, the short hop attack macro, and the inability to cross-up on shields with grounded attacks. The changed perfect shield was also introduced to relieve pressure, and dodges are generally just as strong or stronger than before unstaled.

It also “makes sense” in some insane way because it makes safety on shield rely entirely on endlag. It fits perfectly with the rather childish and severely outdated impression Sakurai has about how Fighting games work.
 
Last edited:

J0eyboi

Smash Ace
Joined
Dec 28, 2017
Messages
534
#51
The Zelda up against Ike was Treehouse staff, right?
That was ZeRo and Leffen.

In a game where pressure would be higher than ever before, making shields that much better would make sense coming from the mind that thought Brawl hitstun canceling was a good idea.
First of all, shield pressure won't be as strong as it was in Melee. Second of all, if the goal was to make shields stronger, he wouldn't do it in a weird, archaic way that's hard for the average player to even understand and actively undermined the concept of shieldstun. He'd just lower shieldstun or remove it from the game entirely. We know this because that's exactly what he did with Brawl.

I have a hard time believing it’s a bug simply because A: It exists here in the game
That makes no sense but ok

B: I don’t see how or why it would be introduced because shield mechanics aren’t that much different than before.
As a programmer, I can think of a huge number of ways this could've accidentally happened. It's very possible.

The changed perfect shield was also introduced to relieve pressure, and dodges are generally just as strong or stronger than before unstaled.
Afaik only one roll in the entire game has had its unstaled frame data buffed, and it's Samus' forward roll.

It also “makes sense” in some insane way because it makes safety on shield rely entirely on endlag. It fits perfectly with the rather childish and severely outdated impression Sakurai has about how Fighting games work.
So ignoring that your view of Sakurai's views on fighting games is outdated, why wouldn't he just remove shieldstun? It's a much simpler solution.
 
Joined
Jan 28, 2014
Messages
7,809
Location
Buried under 990+ weapons
3DS FC
1504-5709-4054
#52
If we had Smash 4 shields and Ultimate landing lag (both more shieldstun and less endlag than Melee on average?) that would be true, correct?

Jumping out of shieldstun would be a pretty big and noticeable bug to just leave in. If anybody ever tried to buffer a jump out of shieldstun it would be immediately obvious.

I would like some explanations, because I dabble in code and stuff too and can’t imagine how it would come up.

Neutral air dodges and spot dodges are much stronger than they were in 4.

Why wouldn’t Sakurai just cut landing lag in 64/Melee if it’s that much simpler? Why would Sakurai make people mash even harder for things like Luigi’s Dspecial and Bowser’s Uspecial? Why would Pikachu’s Side B start stronger with a Smash input? Why make it harder to Uthrow/Dthrow if a character has a Kirby Uthrow? Why why why does basically anything in Smash work the way it does instead of something simpler and easier? No ****ing clue. This at least has a simple explanation of being an “advanced technique” like L-canceling and mashing for recovery.
 
Joined
Apr 13, 2014
Messages
1,636
Location
Santa Ana, CA
#53
Jumping out of shieldstun would be a pretty big and noticeable bug to just leave in. If anybody ever tried to buffer a jump out of shieldstun it would be immediately obvious.
As far as I know, nobody outside this thread is aware of the shield jump. Not even the several Smash 4 veterans that got several hours with the game. So I wouldn't call it "noticeable". It wouldn't be immediately obvious to somebody because they'd probably chalk up their faster responsiveness after blocking to the shorter jumpsquat and lower shieldstun. It's not a complete answer, but it's a plausible answer that nobody would take issue with.

Let me bring up a pertinent example of how a big bug can get through, regarding shields interestingly enough. In Brawl there was a bug where moves with more hitlag (electric attacks, Marth's tipper, etc.) forced more shieldlag on the user, but not a corresponding amount of shieldlag to the defender. So those attacks were arbitrarily made less safe on shield than they were supposed to be. Granted, nothing was very safe on shield in the first place so it didn't impact the game's competitive play much outside of Marth mains trying very hard to never tip opponents on the ground. But this bug actually made it all the way into Smash 4. It wasn't patched out until 1.11, a whole year after game release.

And as far as I can tell, the most likely explanation for why they finally noticed the bug is because Ryu was very new and there are hitlag modifers on all of his attacks. Including his Strong F-Tilt which was supposed to be a block string. So Patch 1.11's changes (equalized shieldlag, and increased shieldstun) may all have been done just to have Ryu's collarbone breaker do its job. It's dumb, but the changes were extremely helpful for toning down the effectiveness of blocking. This shield jump thing can very well be there on release, and may never get patched, but that's not going to stop me from calling it a bug and drawing attention to it.
 
Last edited:

J0eyboi

Smash Ace
Joined
Dec 28, 2017
Messages
534
#54
If we had Smash 4 shields and Ultimate landing lag (both more shieldstun and less endlag than Melee on average?) that would be true, correct?
There would be more shieldstun than Melee, though our current numbers indicate that shieldstun is lower than Smash 4, but the landing lag is about the same. Additionally, the shorter shield drop time, lower pushback, and floatier physics will make pressure weaker overall.

I would like some explanations, because I dabble in code and stuff too and can’t imagine how it would come up.
It depends a lot on how exactly the game's shield mechanics are programmed, which isn't information I'm privy to, but my guess is that they intended to make you able to jump cancel after a successful parry and accidentally made it so you can always jump cancel shields due to an oversight.

Neutral air dodges and spot dodges are much stronger than they were in 4.
Spot dodges sure, but where are you getting that neutral airdodges are stronger? I know they have less landing lag, but besides that I can't find a single source for their FAF, and they're now limited to once per jump unless you get hit, which is a hard nerf.

Why wouldn’t Sakurai just cut landing lag in 64/Melee if it’s that much simpler?
L- and Z-cancelling were meant to be advanced techniques that rewarded accurate inputs and higher technical skill. Sakurai stopped including mechanics like them from Brawl onwards because he believes they cause to much of a gap between player skill levels.

Why would Sakurai make people mash even harder for things like Luigi’s Dspecial and Bowser’s Uspecial?
Because casual players are likely to try to mash while recovering anyway.

Why would Pikachu’s Side B start stronger with a Smash input?
Basically the same as above.

Why make it harder to Uthrow/Dthrow if a character has a Kirby Uthrow?
I have no clue. I don't even know what you're talking about.
 
Last edited:
Joined
Jan 28, 2014
Messages
7,809
Location
Buried under 990+ weapons
3DS FC
1504-5709-4054
#55
As far as I know, nobody outside this thread is aware of the shield jump. Not even the several Smash 4 veterans that got several hours with the game. So I wouldn't call it "noticeable". It wouldn't be immediately obvious to somebody because they'd probably chalk up their faster responsiveness after blocking to the shorter jumpsquat and lower shieldstun. It's not a complete answer, but it's a plausible answer that nobody would take issue with.

Let me bring up a pertinent example of how a big bug can get through, regarding shields interestingly enough. In Brawl there was a bug where moves with more hitlag (electric attacks, Marth's tipper, etc.) forced more shieldlag on the user, but not a corresponding amount of shieldlag to the defender. So those attacks were arbitrarily made less safe on shield than they were supposed to be. Granted, nothing was very safe on shield in the first place so it didn't impact the game's competitive play much outside of Marth mains trying very hard to never tip opponents on the ground. But this bug actually made it all the way into Smash 4. It wasn't patched out until 1.11, a whole year after game release.

And as far as I can tell, the most likely explanation for why they finally noticed the bug is because Ryu was very new and there are hitlag modifers on all of his attacks. Including his Strong F-Tilt which was supposed to be a block string. So Patch 1.11's changes (equalized shieldlag, and increased shieldstun) may all have been done just to have Ryu's collarbone breaker do its job. It's dumb, but the changes were extremely helpful for toning down the effectiveness of blocking. This shield jump thing can very well be there on release, and may never get patched, but that's not going to stop me from calling it a bug and drawing attention to it.
I’ve seen a certain Debby downer repeat this information on a couple of Smash related places, and occasionally I’ll see other people say things like “too bad it doesn’t matter because nothing will be safe” and the like. It’s getting around.

This would be highly suspect to development staff, especially if they cared at all about testing things like how safe moves are. Then again they might just not care, it’s hard to tell. If one person in development ever decided to count frames it would immediately be obvious.

As far as our best and brightest missing things... Well they’ve definitely missed things. ZSS had a Peach style float in one Smash 4 demo and Zero didn’t notice that. Bigger things have slipped through the cracks. Absolutely gigantic things have been missed by all parties in other games still... But this is a universal mechanic, just short of like Yoshi or something. The severe amount of neglect required for this to make it in to a demo build made at the tail end of development is absurd.

I’m not entirely sure the hitstun modifier thing was a bug. Note that it did make it to Ryu’s release, and could of been corrected instead of changing everything. Marth was still kinda not good at that point so it’s not like it would of caused a problem.

You can call attention to it all you want but I still doubt it’s a bug. I suppose we’ll see soon enough, anyway.

There would be more shieldstun than Melee, though our current numbers indicate that shieldstun is lower than Smash 4, but the landing lag is about the same. Additionally, the shorter shield drop time, lower pushback, and floatier physics will make pressure weaker overall.



It depends a lot on how exactly the game's shield mechanics are programmed, which isn't information I'm privy to, but my guess is that they intended to make you able to jump cancel after a successful parry and accidentally made it so you can always jump cancel shields due to an oversight.



Spot dodges sure, but where are you getting that neutral airdodges are stronger? I know they have less landing lag, but besides that I can't find a single source for their FAF, and they're now limited to once per jump unless you get hit, which is a hard nerf.



L- and Z-cancelling were meant to be advanced techniques that rewarded accurate inputs and higher technical skill. Sakurai stopped including mechanics like them from Brawl onwards because he believes they cause to much of a gap between player skill levels.



Because casual players are likely to try to mash while recovering anyway.



Basically the same as above.



I have no clue. I don't even know what you're talking about.
Hm, regardless, it’s something I could see Sakurai thinking is a problem at the beginning of development anyway.

See, I doubt that. They likely started with Smash 4 shields, and added the new parry very early on. You wouldn’t change such a thing most of the way through development. And if they wanted the parry to be jump cancelable then they’d add a flag to that and not other shielding animations.

Less landing lag, and significantly too. If you only need one, and that’s often all it takes, then it’s most certainly stronger.

The mashing is significantly harder than L-canceling ever was. It’s stupidly hard, and even pros have trouble doing it consistently without like weird B-stick shenanigans or something. It’s incredibly casual unfriendly.

In order to input a throw... Imagine an analogue stick viewer. In the center is a square, and touching the edges or going beyond makes you throw. That’s most characters. For characters with those particular types of throws, it’s a more narrow rectangle for whatever reason. It’s much easier to misinput, and that’s by design. And for no real reason either. Kirby and MK’s Uthrows aren’t anything special and Zard’s didn’t kill to begin with (IIRC). It’s a Sakurai-ism, I guess. A weird mechanic that exists to exist.
 

J0eyboi

Smash Ace
Joined
Dec 28, 2017
Messages
534
#56
This would be highly suspect to development staff, especially if they cared at all about testing things like how safe moves are. Then again they might just not care, it’s hard to tell. If one person in development ever decided to count frames it would immediately be obvious.
Why would they count frames? They're making the game. Frame counting has potential for human error, they could just do the math.

The severe amount of neglect required for this to make it in to a demo build made at the tail end of development is absurd.
You realize that for this to be noticed, someone would have to

1) input a jump right as a move hits their shield, which isn't very easy if you aren't expecting a move to hit your shield. You'd basically have to be testing for something exactly like this, especially if shieldstun is lower as it seems to be.
2) Notice that the shieldstun is lower than it should be, which basically requires you to frame count, something they have no reason to do
3) Test it enough to figure out what's happening

Literally the only reason we can be sure it's a thing is because Leffen just so happened to jump out of shieldstun from Ike's Nair significantly before he should have been able to. As far as I know that's the only guaranteed instance of it we've seen happen so far, and you'd pretty much never even notice anything is off if you're watching or playing at normal speed. Additionally, Sakurai has said in an article about hitlag that I'm too lazy to find that he doesn't do a lot of playtesting (which explains a lot, when you think about it) so the chances of them finding a bug like this are fairly minimal.

I’m not entirely sure the hitstun modifier thing was a bug. Note that it did make it to Ryu’s release, and could of been corrected instead of changing everything. Marth was still kinda not good at that point so it’s not like it would of caused a problem.
First of all, could've and would've.

Second of all, if it were intentional but they didn't want it to happen with Ryu, why didn't they just fix it for Ryu?

Third of all, Marth is not the only character whom this affected. A lot of characters have moves with hitlag modifiers or the electric property.

See, I doubt that. They likely started with Smash 4 shields, and added the new parry very early on. You wouldn’t change such a thing most of the way through development. And if they wanted the parry to be jump cancelable then they’d add a flag to that and not other shielding animations.
Again, this is where that whole "I don't know how the game was programmed" thing comes into play. I can think of ways this could happen, but I don't know how the game handles animations or parrying. All I know is that Fox's shine is jump-cancellable starting on the first frame where it reflects projectiles in Melee, and Fox's shine is jump-cancellable after it reflects a projectile in Smash 4.

The mashing is significantly harder than L-canceling ever was. It’s stupidly hard, and even pros have trouble doing it consistently without like weird B-stick shenanigans or something. It’s incredibly casual unfriendly.
It's not supposed to be easy. It's supposed to be intuitive. It rewards you for doing the thing you were going to do anyway. How much it rewards you is irrelevant, so long as it's noticable.
 
Last edited:
Joined
Apr 13, 2014
Messages
1,636
Location
Santa Ana, CA
#58
The spreadsheet currently has ground Darkest Lariat as coming out on frame 9, but it actually comes out frame 5. Also Revenge activates as early as frame 3.
Yeah I haven't finished parsing through much of the new footage yet and knew I'd have to keep an eye on his neutral B.

Regarding these spreadsheets, I won't be updating them very much moving forward. New footage that has been coming out is very definitively of a new build of the game and I don't want information to conflict. I've already seen conflicting information with the Chrom/Dark Samus match posted a while back, where Chrom's F-Smash comes out faster than Roy's in previous footage, and I assume they have the same frame data, at least for that basic move. These will still be "E3 demo" numbers and they served their purpose of showing us what to expect out of balance updates in Ultimate. We got way more footage and events with this build than I was expecting, so the idea of making new charts for new demos feels less essential than it did months ago. I'll still update some stuff regarding previously nonplayable characters, but I don't have the wealth of time I did have when I started the charts. Best to wait for release where I will make all the time I need for the final roster.
 

J0eyboi

Smash Ace
Joined
Dec 28, 2017
Messages
534
#59
Not quite sure where to put this, but we've talked about shield jank here in the past, so I'd like to draw attention to whatever the **** happened here.

Fully charged Roy Fsmash should've broken his shield, but instead it did nothing. No shield damage, no shieldstun, it's not even Richter jumping out.
 
Last edited:
Top