Data Smash Ultimate Frame Data Patch 6.0 (Complete)

Joined
Apr 13, 2014
Messages
1,675
Location
Santa Ana, CA
#1
Hey everyone, I decided to manually test and write up the frame data for all the characters in the game. I did what I could to get some data from footage of the E3 demo, but having the game in my hands allows me to get more data, and in more accurate ways. This game actually has some new features to make the process far quicker than in Smash 4. Most of the data was initially collected from version 1.1 but is updated with the release of new patches and characters. The goal of having these charts so early was to be prepared for the first balance patch. Smash 4's lack of patch notes really had us struggling for the first few months because nobody had the numbers to compare to until about a year after release.

Please let me know if you have any questions or want clarifications. Also you are free to use or repurpose this data as you like. No need to ask first. Of course I will also add new charts for any newly released characters. And the first link of this post will always be the current patch of the game. I make copies of the previous patches which you can see below:

Previous versions:
E3 Demo
Patch 1.1/1.2
Patch 2.0
Patch 3.0
Patch 3.1
Patch 4.0/5.0
 
Last edited:

YashiWins

Smash Rookie
Joined
Feb 2, 2015
Messages
8
Switch FC
SW-4650-5745-8993
#2
Hey everyone, I decided to manually test and write up the frame data for all the characters in the game. I did what I could to get some data from footage of the E3 demo, but having the game in my hands allows me to get more data, and in more accurate ways. This game actually has some new features to make the process far quicker than in Smash 4, so I hope to mop up this massive roster within the next few weeks. Particularly before any new balance updates so that we can confirm changes accurately. As far as the order, I am strongly considering doing the non-echo newcomers next before resuming numerical order as defined by the game. Echo fighters have last priority.

Please don't request I do this or that character next. But let me know if you have any questions or want clarifications. Also you are free to use or repurpose this data as you like. No need to ask first.

Hi, I just wanted to say thank you for undertaking this work. I was able to look at the Yoshi data and found it very helpful. Happy investigating!
 
Joined
Apr 13, 2014
Messages
1,675
Location
Santa Ana, CA
#3
So with a new patch coming next week, I think I'll have to keep my game in patch 1.1 both for the sake of these charts pertaining to the same version of the game and for the community patch notes I expect we'll be stuck doing. I know they claim on twitter there will be detailed patch notes, but I'm going to be skeptical in order to be safe. Nintendo has done official patch notes for numerous recent games since Smash 4 (ARMS, Pokken, both Splatoons) that go over explicit balance decisions, but some of them seemed rife with misinformation, incomplete information, or poor wording/translations.
 
Joined
Apr 13, 2014
Messages
1,675
Location
Santa Ana, CA
#5
Zapp, I can volunteer to help with frame data.

How do you find active frames? The first active frame seems easy, but frames after that?
I guess if you wanted to then you would need to attack somebody while their invulnerable from a dodge. Like, if you wanted to confirm Mario's jab 1 still hits on frame 3, and had a character whose spot dodge invulnerability ends on 16, you'd need to time Mario's Jab input on frame 13 so you hit them on 16.

But I have no intention of testing each active hit frame at this time. There are so many characters in the game and each one takes hours already to get the data points I'm looking for. Dodges also stale which forces you to wait and wonder if your test was late enough for it to be fresh again. Once the roster is complete with their move data, I could go back and add new data points. I was thinking shield stun. At least until we can lock down a proper formula for that.
 
Last edited:
Joined
Apr 13, 2014
Messages
1,675
Location
Santa Ana, CA
#7
You can just turn on invuln display to make sure you didn't mess up dodge staling.
Oh now this is a neat feature. They cooked up a lot of tools in this training mode. It even shows limb intangibility, the new invincibility on throws. Can't seem to get characters to turn blue when they should be on invisible (warp) recoveries or Mewtwo's air dodge, but for general dodges it's perfect.
 

jonnobigz

Smash Apprentice
Joined
Aug 8, 2005
Messages
90
Location
Fremont, CA
#8
I guess if you wanted to then you would need to attack somebody while their invulnerable from a dodge. Like, if you wanted to confirm Mario's jab 1 still hits on frame 3, and had a character whose spot dodge invulnerability ends on 16, you'd need to time Mario's Jab input on frame 13 so you hit them on 16.

But I have no intention of testing each active hit frame at this time. There are so many characters in the game and each one takes hours already to get the data points I'm looking for. Dodges also stale which forces you to wait and wonder if your test was late enough for it to be fresh again. Once the roster is complete with their move data, I could go back and add new data points. I was thinking shield stun. At least until we can lock down a proper formula for that.
Might be easier to test on a character with a move with early, short invulnerability. I'll think of something.
 

Ajani

Smash Apprentice
Joined
Dec 14, 2018
Messages
85
#9
Im looking at doing some Frame data analysis for Lucario. As this is my first time, does this sound right to you guys?


Move..................Frames..................Upstart

F-Smash.............65-70..................15


U-Smash.............60........................1-8 (dependent positioning)


D-Smash.............62........................16



I found each position from the start of the position to when he returned to regular stance. Sorry for the poor formatting, I don't know how to make tables.
 

Nysyr

Smash Journeyman
Joined
Nov 5, 2014
Messages
288
#10
Im looking at doing some Frame data analysis for Lucario. As this is my first time, does this sound right to you guys?


Move..................Frames..................Upstart

F-Smash.............65-70..................15


U-Smash.............60........................1-8 (dependent positioning)


D-Smash.............62........................16



I found each position from the start of the position to when he returned to regular stance. Sorry for the poor formatting, I don't know how to make tables.
Looks way off from my testing. We have a frame advance tool in training for testing. Use shield to find your FAF.

Lucarios Fmash is F19, FAF 54, with like 11 frames of hitlag at 0% on mario
 

DavemanCozy

Smash Photographer
Joined
May 16, 2013
Messages
1,693
Location
London, ON
3DS FC
0216-1810-7681
NNID
CavemanCossy
#11
:ultsnake: Snake notes to add:

Dash attack has intangibility on his head and arms during frames 5-12 (i.e. through the entire duration of the hitbox)

Nikita Side-b grants Snake intangibility for 4 frames right when he fires the missile.
 
Joined
Oct 12, 2016
Messages
49
NNID
josiahs13
#12
So I was looking at the Bird's frame data to see if the numbers I got matched. Mostly they do, but I know that Falco's utilt's first hitbox is active frame 3, not frame 5. I dunno if this is a fluke or how many other things might be in the spreadsheet like this. I know there's a ludicrous amount of data, but it might be good to review or have it checked.
 

Ajani

Smash Apprentice
Joined
Dec 14, 2018
Messages
85
#13
Looks way off from my testing. We have a frame advance tool in training for testing. Use shield to find your FAF.

Lucarios Fmash is F19, FAF 54, with like 11 frames of hitlag at 0% on mario
Thank you

I agree my numbers were off. Do you have the Lucario frame analysis available for viewing?
 
Joined
Apr 13, 2014
Messages
1,675
Location
Santa Ana, CA
#14
:ultsnake: Snake notes to add:

Dash attack has intangibility on his head and arms during frames 5-12 (i.e. through the entire duration of the hitbox)

Nikita Side-b grants Snake intangibility for 4 frames right when he fires the missile.
Yeah you know I'd better add that sort of information now that it's so easy to check. I went back into all the previous charts and updated them. But for Snake's Nikita missile, I could't find any I-frames on the move. I attacked him the frame before he fires, the frame he does fire, the frame after. Nothing.

So I was looking at the Bird's frame data to see if the numbers I got matched. Mostly they do, but I know that Falco's utilt's first hitbox is active frame 3, not frame 5. I dunno if this is a fluke or how many other things might be in the spreadsheet like this. I know there's a ludicrous amount of data, but it might be good to review or have it checked.
I don't doubt the potential for mistakes, though I've been doing this sort of thing for years. But I'm afraid his Utilt hits on 5, not 3.
 

DavemanCozy

Smash Photographer
Joined
May 16, 2013
Messages
1,693
Location
London, ON
3DS FC
0216-1810-7681
NNID
CavemanCossy
#15
for Snake's Nikita missile, I could't find any I-frames on the move. I attacked him the frame before he fires, the frame he does fire, the frame after. Nothing.
that's odd, when I did training mode his body was glowing blue, which means he's intangible. Thanks for double checking, I'll check this when I get home (maybe my eyes are playing tricks on me).
 
Joined
Oct 12, 2016
Messages
49
NNID
josiahs13
#16
I don't doubt the potential for mistakes, though I've been doing this sort of thing for years. But I'm afraid his Utilt hits on 5, not 3.
By jingo, it does. What do you know. I could swear it WAS frame 3, anyway. I actually didn't download the day 1 update (because I had automatic downloads turned off and I was feeling a touch impatient to play the game), so this could mean they actually made balance changes from 1.0 to 1.1. Interesting...
 

san.

1/Sympathy = Divide By Zero
Moderator
Joined
Oct 26, 2008
Messages
5,632
Location
Rochester, NY
3DS FC
4957-2846-2924
NNID
Sansoldier
#17
Thanks for the frame data on :ultike:

A few observations:
  • UpB, Aether, has super armor on frame 5 when grounded, and frame 15 when used in the air.
  • SideB, Quick Draw's attack, does not seem to have the reduced FAF if he does not touch the ground, such as when it's used offstage. It's still reduced if used in the air and he lands before the endlag finishes. I tested by going frame by frame and looking for upB, so I haven't tried this with double jump or air dodge.
That's it so far, everything looks accurate for him.
 
Joined
Apr 13, 2014
Messages
1,675
Location
Santa Ana, CA
#18
By jingo, it does. What do you know. I could swear it WAS frame 3, anyway. I actually didn't download the day 1 update (because I had automatic downloads turned off and I was feeling a touch impatient to play the game), so this could mean they actually made balance changes from 1.0 to 1.1. Interesting...
Can't imagine they made it frame 3 just to revert it to Smash 4's startup later. In fact I have a source from the latest demo before release (which may or may not have been the same build as version 1.0) that also has it on frame 5.

Thanks for the frame data on :ultike:

A few observations:
  • UpB, Aether, has super armor on frame 5 when grounded, and frame 15 when used in the air.
  • SideB, Quick Draw's attack, does not seem to have the reduced FAF if he does not touch the ground, such as when it's used offstage. It's still reduced if used in the air and he lands before the endlag finishes. I tested by going frame by frame and looking for upB, so I haven't tried this with double jump or air dodge.
That's it so far, everything looks accurate for him.
I've got Aether's super armor values on the ground, but I didn't anticipate they'd change for the air version. The window is frame 15-35 in the air. That's interesting about Quick Draw as well. I knew about the recovery mechanic in Smash 4 to avoid getting gimped, but I assumed it would have been updated to the new duration. I've added a note about this to the charts.

Say, you're a mod. Can you put an Important or Data tag on this thread? I'm sure this isn't the only source for frame data on the web, but as far as I know it's the only one with intent to collect info on all the characters. Plus I'll always be around if something needs testing or clarification.
 

san.

1/Sympathy = Divide By Zero
Moderator
Joined
Oct 26, 2008
Messages
5,632
Location
Rochester, NY
3DS FC
4957-2846-2924
NNID
Sansoldier
#19
No, I can't. I only have limited powers for certain character boards.

I fortunately found out in friendlies when I misjudged a recovery thinking the side-B would recover faster than it did due to the lag reduction. You can check to verify what I said matches if you haven't already.
 

BJN39

Smash Data Ranger
Moderator
Joined
Aug 14, 2012
Messages
5,036
Location
The Zelda Boards
#20
Saw this on Zelda's data under Nayru:
"Hits begin 2 frames earlier. Total frames now 57 from 59. Invulnerable on frame
4-29 from 5-15. Reflects on 4-41 from 6-43."

The move is actually invulnerable 4-13 and to my understanding reflects frame 5-41 changed from 5-43.

In addition, Jab transitions to rapid earliest frame 9, hits frame 3, and all hits are 1F apart. (8 hits minimum, so 3,5,7,9,11,13,15,17) It was a pain to test but I can verify the gaps between hits. Up smash also uses full length hitboxes and rehit rates (frames 9-23, 25-32, 34. rehit rate of 4. Noticed weird things happening when testing and found this out!)

I guess if you wanted to then you would need to attack somebody while their invulnerable from a dodge. Like, if you wanted to confirm Mario's jab 1 still hits on frame 3, and had a character whose spot dodge invulnerability ends on 16, you'd need to time Mario's Jab input on frame 13 so you hit them on 16.
The way I've tested hit frames before is by using a jumping CPU in training and trying various different timings and distances to see what hits and when. It works 90% of the time. Some really obscure hitboxes would be nigh impossible to truly test tho (Like if Zelda's kick sweetspots are longer than 1F.)
 
Last edited:
Joined
Apr 13, 2014
Messages
1,675
Location
Santa Ana, CA
#21
Saw this on Zelda's data under Nayru:
"Hits begin 2 frames earlier. Total frames now 57 from 59. Invulnerable on frame
4-29 from 5-15. Reflects on 4-41 from 6-43."

The move is actually invulnerable 4-13 and to my understanding reflects frame 5-41 changed from 5-43.

In addition, Jab transitions to rapid earliest frame 9, hits frame 3, and all hits are 1F apart. (8 hits minimum, so 3,5,7,9,11,13,15,17) It was a pain to test but I can verify the gaps between hits. Up smash also uses full length hitboxes and rehit rates (frames 9-23, 25-32, 34. rehit rate of 4. Noticed weird things happening when testing and found this out!)



The way I've tested hit frames before is by using a jumping CPU in training and trying various different timings and distances to see what hits and when. It works 90% of the time. Some really obscure hitboxes would be nigh impossible to truly test tho (Like if Zelda's kick sweetspots are longer than 1F.)
Thanks for the heads up on rapid jab. I see now the frame of animation inbetween hits. On Usmash I guess I wanted to see it as the same attack since it ends on the same hit frame as before. And for Nayru's love. I thought I had managed to attack through her at the end of the move, but cannot reproduce it. Glad to hear I at least got the final reflect frame correct, since it's the hardest to confirm what with you not being able to sit next to her with a link arrow ready to fly, but instead be finely spaced away from her final hitbox.
 
Joined
Apr 13, 2014
Messages
1,675
Location
Santa Ana, CA
#22
Well it looks like I'm finished with the frame data. It took somewhere over 100 hours to count out all these numbers, and I hope people can find use for all of it. Especially for the first balance patch which I anticipate could arrive as early as January. Having these charts to refer back to should help quite a bit for confirming changes with precise numbers.

I've coupled all the echo fighters with their original counterparts, with the exception of Ken. I think that Ken not only has a notable number of unique moves and properities, but also it would look especially ugly to stick his moves in with Ryus given the format of the charts. It was just easier to have a separate chart. I still don't know what I should do as far as adding data points to the charts. Shieldstun and block advantage are how I got my start with frame counting, so maybe that. If I do do shieldstun, Chrom, Lucina, and I suppose Dark Pit will need to be separated since they have obvious damage differences. I've considered the possibility of Damage and knockback just to have more comprehensive lists of how these moves have changed, but I have no way to get knockback in units or reasonably compare it to the last game. Even if I lined up kill percents between games all I could say is the knockback is improved in some way, not its growth or base.

Anyway, that's all that's on my mind. Happy holidays, Smashboards.
 
Last edited:

Ajani

Smash Apprentice
Joined
Dec 14, 2018
Messages
85
#23
Well it looks like I'm finished with the frame data. It took somewhere over 100 hours to count out all these numbers, and I hope people can find use for all of it. Especially for the first balance patch which I anticipate could arrive as early as January. Having these charts to refer back to should help quite a bit for confirming changes with precise numbers.

I've coupled all the echo fighters with their original counterparts, with the exception of Ken. I think that Ken not only has a notable number of unique moves and properities, but also it would look especially ugly to stick his moves in with Ryus given the format of the charts. It was just easier to have a separate chart. I still don't know what I should do as far as adding data points to the charts. Shieldstun and block advantage are how I got my start with frame counting, so maybe that. If I do do shieldstun, Chrom, Lucina, and I suppose Dark Pit will need to be separated since they have obvious damage differences. I've considered the possibility of Damage and knockback just to have more comprehensive lists of how these moves have changed, but I have no way to get knockback in units or reasonably compare it to the last game. Even if I lined up kill percents between games all I could say is the knockback is improved in some way, not its growth or base.

Anyway, that's all that's on my mind. Happy holidays, Smashboards.
From me atleast, Thank you. Your dedication should not go unnoticed.
 

Blue Banana

Smash Journeyman
Joined
Oct 11, 2014
Messages
409
3DS FC
4038-6328-0283
NNID
2ndDerivative
#24
Hi, I noticed there are some numbers in Olimar's frame data that seem to contradict with the data that I have.
- According to the data from KuroganeHammer, Olimar's Dair in Smash 4 starts becoming active on frame 9, not frame 8. This hasn't changed in Ultimate.
- The total duration for Forward Throw, Back Throw, and Up Throw are the same as Smash 4, according to the data that ScherzoGavotte ScherzoGavotte obtained.
- Down Throw's endlag has been decreased by 4 frames, based on Scherzo's data (Total duration: 45 -> 41).

Everything else seems to be correct.
 
Last edited:
Joined
Apr 13, 2014
Messages
1,675
Location
Santa Ana, CA
#25
Hi, I noticed there are some numbers in Olimar's frame data that seem to contradict with the data that I have.
- According to the data from KuroganeHammer, Olimar's Dair in Smash 4 starts becoming active on frame 9, not frame 8. This hasn't changed in Ultimate.
- The total duration for Forward Throw, Back Throw, and Up Throw are the same as Smash 4, according to the data that ScherzoGavotte ScherzoGavotte obtained.
- Down Throw's endlag has been decreased by 4 frames, based on Scherzo's data (Total duration: 45 -> 41).

Everything else seems to be correct.
Oh, weird. Maybe I made a typo on D-air's entry when I drew up Olimar's charts for Smash 4 years ago. Well, I don't have any smash 4 footage on hand to fact check, so I'll remove that suspected difference. As for the throws, that chart should suffice in what I'm referring to. Olimar's throws have the same animation length, but the startup has been pushed back, meaning endlag has been reduced in order to match the old total frames. This is a very common change with throws in Ultimate from Smash 4 that now have new animations. Furthermore, my findings years ago had identical numbers for his Dthrow in Smash 4 as I had found in Ultimate, so I judge there must be no change.
 
Joined
Apr 13, 2014
Messages
1,675
Location
Santa Ana, CA
#27
Is there data on pivot grabs in the charts?
Nope. I got frustrated because we can't run through people anymore. Thus removing the obvious, consistent method to test the startup of those grabs as easily as testing any other move. Not to mention removing the obvious, useful application of those grabs in the first place unless your grab is a tether grab or has incredible range like Bowser's. I suppose I could add them later. Just control the the victim and have them dash into the characters as they perform a pivot grab.
 

san.

1/Sympathy = Divide By Zero
Moderator
Joined
Oct 26, 2008
Messages
5,632
Location
Rochester, NY
3DS FC
4957-2846-2924
NNID
Sansoldier
#28
Cool, thanks. To help with this, as you plug in your controller for player 2, hold the opposite direction on the control stick so that once the stick resets to neutral, it holds the intended direction.
 

Shaya

   「chase you」 
Administrator
GRimer
Rankings Team
Joined
Jun 8, 2007
Messages
27,658
Location
/人◕‿‿◕人\ FABULOUS Max!
NNID
ShayaJP
#29
Zero Suits noticed that the spreadsheet has invincibility on up smash frames 8-14?
Wasn't able to observe any limbs going intangible in training mode, and ZSS frame 1 jab will hit ZSS on frames 8 and 9, at the least.

Would be really cool if she was intangible during up smashes start up, but currently don't think so.
 
Joined
Apr 13, 2014
Messages
1,675
Location
Santa Ana, CA
#30
Zero Suits noticed that the spreadsheet has invincibility on up smash frames 8-14?
Wasn't able to observe any limbs going intangible in training mode, and ZSS frame 1 jab will hit ZSS on frames 8 and 9, at the least.

Would be really cool if she was intangible during up smashes start up, but currently don't think so.
Oh weird, that information was meant for Wario. Must have thumbed to the wrong chart when adding that. It's fixed now.
 
Joined
Apr 13, 2014
Messages
1,675
Location
Santa Ana, CA
#35
The first hitbox has always been quite potato, I cant test right now but you might need a jumping enemy above Lucario for it.
That's what I figured when I discovered it. Tried a few methods to bury Lucario's palm inside somebody's hitbox or shield but it always on 7.
No its 7. I was doing an incorrect input. Honestly, Im just shocked at the upstart. oh well.
That reminds me to remind folks that I just count the numbers. I don't decide the changes, obvious as they might be. What shocks me more than anything is how half the dash attacks in the previous game got a considerable boost in some department - since these moves now compete with every other move on the ground for utility, but Marth was left out. And he had pretty definitively the worst one.

Edit: I have also now added Pivot Grabs since that was suggested earlier. And Neutral Air Dodges. I'm still considering how Directional Air Dodges ought to look with this format, since they seem to be quite complicated and certainly very time consuming to cover.
 
Last edited:

Thinkaman

Moderator
Moderator
Joined
Aug 26, 2007
Messages
6,060
Location
Madison, WI
3DS FC
1504-5749-3616
NNID
Thinkaman
#36
You might want to express Spot Dodge total frames as X/Y alternative values, such as 21/25 for Mario.

(Attacks and specials can be done exactly 4 frames sooner out of spot dodge for all characters than grab, jump, and defensive options; including both values makes the data totally unambiguous as to which you are expressing.)
 
Joined
Apr 13, 2014
Messages
1,675
Location
Santa Ana, CA
#37
You might want to express Spot Dodge total frames as X/Y alternative values, such as 21/25 for Mario.

(Attacks and specials can be done exactly 4 frames sooner out of spot dodge for all characters than grab, jump, and defensive options; including both values makes the data totally unambiguous as to which you are expressing.)
What, I didn't know about this.

Why do I feel like I need a textbook to understand dodging in Ultimate? I better start dumping notes about dodges into the glossary chart. Both for myself and those who are as lost as I am.
 
Joined
Apr 13, 2014
Messages
1,675
Location
Santa Ana, CA
#38
Well all the dodge stuff is finished. I also added more helpful notes to the Glossary page. And I got shield data for Mario just to see how the formatting would look - it looks pretty ugly. Moving forward I don't think I'll collect the numbers for short hop aerials. And I wouldn't expect shield stats to come with the same speed as last month's updates since I won't be spending as much time on these charts. Anyway, here's some observations

  • The 1v1 damage modifier is applied after calculations, thus shieldlag and shieldstun in a 1v1 setting is identical to matches with more than two players
  • Comparing directly to Smash 4, shieldstun for grounded moves appears slightly higher than Smash 4. But aerial shieldstun is drastically reduced. Projectiles may also be slightly reduced but I only have one point of comparison.
  • I don't know why Mario's Dash attack has so much shieldstun. There may be something special about certain attacks. I thought perhaps all dash attacks have some unique formula, but I have not found another one with an abnormal amount of shieldstun
  • No, I have not looked at perfect shields yet.
 

Thinkaman

Moderator
Moderator
Joined
Aug 26, 2007
Messages
6,060
Location
Madison, WI
3DS FC
1504-5749-3616
NNID
Thinkaman
#39
The 1v1 damage bonus seems to be the very last modification made, and is done on the receiving end of the player. Other damage-able entities (like say DH Gunmen) take standard, non-increased damage.
 
Joined
Apr 13, 2014
Messages
1,675
Location
Santa Ana, CA
#40
The 1v1 damage bonus seems to be the very last modification made, and is done on the receiving end of the player. Other damage-able entities (like say DH Gunmen) take standard, non-increased damage.
That's what I figured was the explanation. Like how your KO percents on a raw smash attack won't change on either setting yet the KO percents on a Throw > U-air sort of kill confirm will. Or why Mario's Jab 2 fails to knock back gordos despite dealing 2.0 on 1v1 settings. Or why having knockback or damage based armor does not change what moves you can blast through on either setting. I didn't anticipate they'd do this with the damage multiplier but boy does it make the job of testers easier by not having this additional variable to work around on so many test cases.
 
Top