Pre-release Frame Data – Video Analysis

[Vipermoon]

Using YouTube's frame by frame tool (<>) on Marth and Lucina footage, I counted frames with videos set to HD (60 fps). Allow for an uncertainty of ±1 frame using a method like this.

DB1 frame 9, total 12 (before next input)
DB2 frame 4, total 6
DB3 frame 4, total 6
DB4 frame 7, total 55

DB1 frame 9, total 12
DB2 frame 4, total 6
DB3-D frame 5, total 6
DB4-D frame 7 10 13 16 19, total 75(?)

DB1 frame 9, total 12
DB2-U frame 3, total 6
DB3-U frame 5, total 7
DB4-U frame 6, total ???

Counter frame 6, freeze 21, Attack 4-6

Fsmash frame 10

SB angled upwards, frame 19 (8 frame release)

Jab frame 5, less vertical range (likely to differentiate from Ftilt/Utilt), total 14 (before next input); Jab 2 frame 4 (3 possible), total 28

Uair frame 5

Usmash frame 14, total 59

Go to kuroganehammer.com for Smash 4 Marth's complete frame data.
For Smash 4 Move and Hitbox viewer, go here: https://struz.github.io/smash-move-viewer/#/v1/

Comparison summary:

• DB1 slower startup (+3). Huge nerf, as this is a pivotal out of shield and out of dash option.
• DB2,3,4 much faster startup for remaining DB; but really DB1's startup is all that matters
• DB4-All more end lag due to the faster startup
• DB4-D less frames between hits (-3)
• DB all new animations
• Counter startup same (±0)
• Counter success freeze frame longer (+5)
• Counter's Attack faster startup (-1)
• Fsmash startup same (±0)
• SB startup including charge point and release time same (±0)
• Jab 1 startup same (±0), total before next input faster (-2 vs S4 Marth; -3 vs S4 Lucina)
• Jab 2 startup slower or same (+1 or ±0), total same (+0)
• Uair startup same
• Usmash startup and therefore total laggier (+1)

-----------------------------------------------Update with more frame data-----------------------------------------------------------
Update with aerial/jump/jumpsquat info... (also added to OP)

• Jumpsquat: in the air on frame 4
• Short up air time: 41 frames
• Nair: same animation speed; frame 6 and frame 15; cannot short hop autocancel
• Fair: same animation speed; frame 6; autocancels in short hop; takes soft landing (animation ends before landing)
• Uair: same animation speed; frame 5; autocancels in short hop; takes hard landing (animation still active, landing within autocancel window)

Comparison to Smash 4

• Jumpsquat is two frames faster (even faster than Melee Marth). All characters seem to have a faster jumpsquat in this game.
• Short hop same air time
• Nair, Fair, Uair: no perceived differences (not including damage)

 

[DrRiceBoy]

Damn... I abused the heck out of DB in Smash 4, sad to see it nerfed. Seems like the rest of his moveset is fine though, shouldn't be too big of an issue. Let's hope they buff it again in time for release.

Thanks for your hard work! Very helpful to know. =)

 

[Vipermoon]

Update with aerial/jump/jumpsquat info... (also added to OP)

• Jumpsquat: in the air on frame 4
• Short up air time: 41 frames
• Nair: same animation speed; frame 6 and frame 15; cannot short hop autocancel
• Fair: same animation speed; frame 6; autocancels in short hop; takes soft landing (animation ends before landing)
• Uair: same animation speed; frame 5; autocancels in short hop; takes hard landing (animation still active, landing within autocancel window)

Comparison to Smash 4

• Jumpsquat is two frames faster (one faster than Melee). All characters seem to have a faster jumpsquat in this game.
• Short hop same air time
• Nair, Fair, Uair: no perceived differences (not including damage)

Seeing all of this unchanged means Smash 4 Marth players will not need to adapt to much, helping them be successful at game launch (which means this Marth will get nerfed or at least not buffed LOL).

It is possible more of these will be changed for the final release.

 

[Bowserboy3]

This video might be a good tool Vipermoon - ZeRo uses Marth twice in two different games at the end of the day 2 stream (linked is the first game).

(As an aside, the way ZeRo was playing, especially in game 2, gave me hope that Marth will still be VERY good in this game. The movement was f***ing crisp!)

Also, is it just me, or does Fsmash seem to have a little more reach? Am I the only one thinking this? Definitely not Melee, but looks closer to Brawl IMO.

And actually, what is this frame by frame tool? I wouldn't mind looking at some of this stuff myself tbh.

https://youtu.be/8O-ZUXGcuWU?t=13211

 

[Vipermoon]

This video might be a good tool Vipermoon - ZeRo uses Marth twice in two different games at the end of the day 2 stream (linked is the first game).

(As an aside, the way ZeRo was playing, especially in game 2, gave me hope that Marth will still be VERY good in this game. The movement was f***ing crisp!)

Also, is it just me, or does Fsmash seem to have a little more reach? Am I the only one thinking this? Definitely not Melee, but looks closer to Brawl IMO.

And actually, what is this frame by frame tool? I wouldn't mind looking at some of this stuff myself tbh.

https://youtu.be/8O-ZUXGcuWU?t=13211

I think it's just you. Fsmash's range/reach looks completely identical. And it still only sometimes hits under a battlefield platform (pretty much only reaches when catching landings).
About frame by frame on YouTube, just use the <> keys while paused. Keep in mind not every video or resolution setting will show 60 FPS.

Yeah I was going to check out day 2. So looking at ZeRo's matches and the videos VGBC posted, I can see startup and endlag of every single one of Marth's aerials being identical to Smash 4. Landing lag I've been able to catch is as follows: Nair 8 frames, Bair 10 frames, Dair 13 frames. Fair and Uair are probably in the 8 to 10 range as well.

Damage: Inconsistent. There for certain has to be more than one version of this game running around E3. Sometimes Marth's damage is at Smash 4 level plus buffs. Others it's nerfed damage (think Smash 4 Release Marth). Applies to most if not all characters (for example, Snake Ftilt or Nair damage varying drastically). And I'm only talking about 1 v. 1s, I am aware that 3+ players has different damage.

Also about damage: freshness bonus is higher. Instead of x1.05 (+5%), it's at about x1.07 (+7%).

Also DB hits 2 and 3 have Smash 4 total frames (I previously reported on hit 4), which means they have far more end lag (since startup is faster). Another thing I want to say is characters are falling out of DB even though it does just ~10% fresh. And yet, this move truly looks like it has less range and kill power (and Marth does not step forward as much).

Moving on to hitstun, since hitstun is reduced in Smash 5, and in Smash 4 Marth already barely had his combos/tight windows (like for throws and aerials), we do not have true throw combos and aerial strings (without landing; with landing there's plenty) in Smash 5.

Last thing is Jab 1. While still 25 total frames, it now has zero vertical range (basically frame 7 no longer active and sword trail was tucked in to reflect this) and does extremely little knockback that does not scale with damage. Only enough to connect Jab 2. HUGE nerf to Jab.

I barely have any time to put more than this amount of effort into analysis of this game.

Edit: Also, SB has the same startup/end lag/total frames as Smash 4.

 

[Bowserboy3]

I think it's just you. Fsmash's range/reach looks completely identical. And it still only sometimes hits under a battlefield platform (pretty much only reaches when catching landings).
About frame by frame on YouTube, just use the <> keys while paused. Keep in mind not every video or resolution setting will show 60 FPS.

Yeah I was going to check out day 2. So looking at ZeRo's matches and the videos VGBC posted, I can see startup and endlag of every single one of Marth's aerials being identical to Smash 4. Landing lag I've been able to catch is as follows: Nair 8 frames, Bair 10 frames, Dair 13 frames. Fair and Uair are probably in the 8 to 10 range as well.

Damage: Inconsistent. There for certain has to be more than one version of this game running around E3. Sometimes Marth's damage is at Smash 4 level plus buffs. Others it's nerfed damage (think Smash 4 Release Marth). Applies to most if not all characters (for example, Snake Ftilt or Nair damage varying drastically). And I'm only talking about 1 v. 1s, I am aware that 3+ players has different damage.

Also about damage: freshness bonus is higher. Instead of x1.05 (+5%), it's at about x1.07 (+7%).

Also DB hits 2 and 3 have Smash 4 total frames (I previously reported on hit 4), which means they have far more end lag (since startup is faster). Another thing I want to say is characters are falling out of DB even though it does just ~10% fresh. And yet, this move truly looks like it has less range and kill power (and Marth does not step forward as much).

Moving on to hitstun, since hitstun is reduced in Smash 5, and in Smash 4 Marth already barely had his combos/tight windows (like for throws and aerials), we do not have true throw combos and aerial strings (without landing; with landing there's plenty) in Smash 5.

Last thing is Jab 1. While still 25 total frames, it now has zero vertical range (basically frame 7 no longer active and sword trail was tucked in to reflect this) and does extremely little knockback that does not scale with damage. Only enough to connect Jab 2. HUGE nerf to Jab.

I barely have any time to put more than this amount of effort into analysis of this game.

Edit: Also, SB has the same startup/end lag/total frames as Smash 4.

You are a gent and I salute you. You do too much for us Marth players. It's good to be back seeing you still at it. Been gone for tooooo long LOL!

 

[Vipermoon]

You are a gent and I salute you. You do too much for us Marth players. It's good to be back seeing you still at it. Been gone for tooooo long LOL!

I guess there was nothing else to learn about Smash 4, or about Marth, or anything else to discuss strategy-wise. Plus life gets in the way. Thank you though, I just want to provide Marcina players with as many tools and knowledge as I can.

Next when I get the chance, I'll frame by frame aerial DS so see if it's still intangible frame 1.

 

[Vipermoon]

Aerial DS is still intangible on frame 1, however, there is still a huge nerf. Previously, it you were able to release it, it was guaranteed to not only have a hitbox, but to beat out (initially) other hitboxes. It's because the intangibility went from frames 1-5, with the hitbox out on 5.
In SSBU, intangibility is frames 1-4. Frame 5, the first hitbox frame, is now naked. So a recovering Marth won't be as deadly, and getting out of combos will result in more trades than usual.
DS's landing lag is somewhere between 20-23 frames. Smash 4's is 23.

Dash Attack has a different animation, I think that's obvious. But I found out the hitbox is still frame 13, and it still ends on 49 – which is why it's still a bad dash attack.

Fsmash has the same total frames. Ends on frame 51, same thing.

Dtilt is still frame 7 and still ends on 23. Ftilt still comes out on frame 8, Utilt still frame 6.

All in all, in SSBU E3-version, this Marth is far more similar than he looks. Really, except for a lot of redesigned animations (cool), reduced landing lag, some weaker tippers, and DB speed – it's all carryover.

 

[InfinityCollision]

Airborne on frame 4 (3f jumpsquat) would make his jump startup faster than Melee; Melee/Brawl Marth had 4f jumpsquat.

 

[KuroganeHammer]

thanks for ur hard werk

 

[LancerStaff]

I’m quite confident that freshness is the same... I checked Pit’s damage in a FFA and it worked out to be almost all the same as it was in Smash 4. Go figure lol.

There seems to be inconsistencies with the 1v1 bonus, or maybe we just don’t understand it yet. For example Pit’s pummel seems to be the same in both. I personally recommend using FFA damage as it is “cleaner” though you’re welcome to use what you wish.

I’m interested though, what makes you think there’s two versions?

 

[Vipermoon]

I’m quite confident that freshness is the same... I checked Pit’s damage in a FFA and it worked out to be almost all the same as it was in Smash 4. Go figure lol.

There seems to be inconsistencies with the 1v1 bonus, or maybe we just don’t understand it yet. For example Pit’s pummel seems to be the same in both. I personally recommend using FFA damage as it is “cleaner” though you’re welcome to use what you wish.

I’m interested though, what makes you think there’s two versions?

Actually, Bowserboy figured it out in the Marth general thread. If 1 v1 with Items On, no damage bonus. Damage bonus is only for 1 v 1 No Items [fox only, final destination]
And yeah, considering the FFA/Doubles/Items damage is almost identical to Smash 4 damage, I consider that default as well.

 

[LancerStaff]

That makes sense... Well actually it’s really weird and unintuitive but it makes sense of things lol.

Imagine how much easier our jobs would be if they were a bit more transparent about mechanics and stuff.

 

[Vipermoon]

You take a lot more damage and stocks with items so it makes sense.

 

[LancerStaff]

You do, but it begs more questions. Like, does the stage hazard toggle effect it? Stages outright? Custom equipment? Special Smash?

To me anyway it seems like a hazy place to draw a line. If it was solely determined by player count it would be less confusing.

 

[Bowserboy3]

Vipermoon

Are you seeing any sorts of discrepancies with landing lag from this video compared to your previous records?

https://www.youtube.com/watch?v=7Orq-v4NRrE

Like, let's take Fair for example.

There are times during that video where it looks like Fair is more than 10 frames, but if you take the game 1 winning Fair by Leo, and count the landing lag on that, it's clearly 10 frames.

I am very confused. Maybe I am just bad.

 

[A_Kae]

Vipermoon

Are you seeing any sorts of discrepancies with landing lag from this video compared to your previous records?

https://www.youtube.com/watch?v=7Orq-v4NRrE

Like, let's take Fair for example.

There are times during that video where it looks like Fair is more than 10 frames, but if you take the game 1 winning Fair by Leo, and count the landing lag on that, it's clearly 10 frames.

I am very confused. Maybe I am just bad.

So, looking in to this really quick, the first two fairs used and landed with do seem to having different landing lag. First one looks like 10, second like 14. Other aerials used throughout the video seem to be similarly inconsistent.

Hmmm. More investigation is needed...

Edit: Unless the players aren't acting as fast as possible or there's some weird recording/misc video issue, I can't figure this out. There doesn't seem to be any pattern to this that I can see.

 

[Vipermoon]

Vipermoon

Are you seeing any sorts of discrepancies with landing lag from this video compared to your previous records?

https://www.youtube.com/watch?v=7Orq-v4NRrE

Like, let's take Fair for example.

There are times during that video where it looks like Fair is more than 10 frames, but if you take the game 1 winning Fair by Leo, and count the landing lag on that, it's clearly 10 frames.

I am very confused. Maybe I am just bad.

They're not acting immediately. That's something I had to keep in mind while I tried to get landing lag numbers. It's been happening a lot in the gameplay videos, as people aren't used to the nearly halved landing lag of all the characters. The landing lag animations are almost always longer than the landing lag FAF.

 

[Vipermoon]

FYI, Marth's initial frame data, as accurate as it could be right now:

https://docs.google.com/spreadsheet...t9Fk2zmmSvcEH8rBy6GVAHEtQk/edit#gid=371959282
https://docs.google.com/spreadsheet...S3D1t_ttTracY-SsGXqiChBfPE/edit#gid=721381806

Thanks to M Meshima and Zapp Branniglenn , respectively.