Marth Frame Data

[B!squick]

Is Marth's Dash Attack really so bad that you skipped it entirely? O.o Also, I'm sure this has been asked, but is Marth really invincible during the first hitbox of DS? That's kinda stupid if it is. Makes it better than even Fortress OoS. x_x

 

[Steel]

i only just now realized i forgot to do dash attack. but yeah it's pretty bad. and yes it's invincible on frame 1

 

[Remzi]

i only just now realized i forgot to do dash attack. but yeah it's pretty bad. and yes it's invincible on frame 1

Without reading the post above this one I thought Dash attack had invincibility frames

 

[B!squick]

No no, the first hitbox (my post said frame, but I edited it real quick just after posting which may explain the confusion here). You have the Invincibility frames listed as lasting from frames 1-5 and the first hitbox listed as coming out on frame 5. This would mean that as long as you're in range the first hit is uninteruptable. Always. Ever in life. D:

 

[Emblem Lord]

Yeah.

This is pretty common in other fighters.

Clearly Sakurai took a cue from Street Fighters Shoryuken where the first few frames are invincible.

 

[B!squick]

Dam. Ah well. At least you can ledge cancel Fortress, so in that respect it's better at least.

 

[3xSwords]

Can you get the frame data for the actual strike by counter? Like number of invincibility frames on it while countering and all the other good stuff.

 

[Steel]

smash lab is going to start providing official frame data sometime soon so i'm not going to do anything more for now

 

[B!squick]

smash lab is going to start providing official frame data sometime soon so i'm not going to do anything more for now

Woah, really!? Link plox?

 

[Steel]

It's just a part of our agenda >_> There's no public link for it

 

[Pr0phetic]

Woah, really!? Link plox?

Keywords: sometime SOON.

 

[Shaya]

Counter 'Hit'

(Opponent attack hits on frame '0')
Invincibility: 1-22
Hit: 19-22
Counter Hit Lag: 15 (this applies to physical hits, and doesn't apply to you, only your opponent. This isn't added to the opponent's normal hitlag, so it instead 10 + 15, it's JUST 15)
Hit Lag: 9 (11 tip)
Shield Stun: 13
Hit Stun (air dodge): 15 (14 tip)
Hit Stun (attack): 25
End: 50

Advantage (shield): -29 (-31 tip)
Advantage Hit: -16 (-17 tip)

 

[3xSwords]

<3 you Shaya

 

[Yoshi Kirishima]

Wow, thanks for all this stuff :D Just absolutely great and helpful!

 

[One_With_Sumthing]

A lot of your frame data is ambiguous/wrong...

If IASA of Dtilt is frame 21 (i.e. frame 1 of any other action is 22), then why do you have end frames of many moves 1 frame later? Unless this was an error in your data collection, in this case, there are quite a few holes.

i.e. DB1 'IASA' is 29, frame 1 of any other action can be 30. Fair is 33. etc etc.


Also,
Marth's bair has "Zero lag" from frame 31.

Marth's dair has "Zero Lag" from I presume frame 45/46 (getting people to land on the ground at the right times are irritating).

Marth has 2 frames of landing lag from a short hop.
Marth has 4 frames of landing lag from a full hop.

Marth's jump has 5 frame start up.
Input on frame 5 is ignored.
From frame 1-4:
Any up smash or up b cancels the jump.
Any aerial can be buffered.

Also (lol)
Dolphin Slash has 3 frames hit lag, 10 frames shield stun.
You have frames 1-3 of dolphin slash to do a turn around.


Also, I've already told you that untippered dtilt's frame advantage is -9

what's the landing lag from a second jump?

 

[Dark Sonic]

I can't help but notice that you didn't list dash grab, pivot grab, or dash attack. So I'll list them for you. And you should probably check shieldhitlag for both Marth and the opponent seperately, because apparently some moves cause different shieldhitlags for the attacker and the shielder.

Dash attack
Hit: 13-16
end: 49
shieldstun: 2
shieldhitlag (Marth): 5
shieldhitlag (opponent): 7
tipped shieldstun: 2
tippedshieldhitlag (Marth): 11
tippedshieldhitlag (opponent): 8

Dash grab
hit: 9
end: 39

Pivot grab
hit: 9-10
end: 35

edit: Redid. Don't do frame data at 3 in the morning lol.

 

[Yoshi Kirishima]

Great awsome :D

btw how do you guys figure out all this frame stuff? Do you just go into a video editor and watch it by frames or what? o.0

 

[Dark Sonic]

I used the debug pause code to let me play the game frame by frame. Then I just counted. Shield hitstun was done by counting from when the move hit the shield until there was any sign of movement by the person in question (so, counting until the shielder started sliding for him, and counting until Marth's blade started moving again for Marth.) Durations were calculated by timing the attack to overlap the ending of DDD's spotdodge, which can be hit after frame 20 (so I'd time the attack to have a hitbox out on 21, then bump it backwards 1 frame until it no longer hits).

Also, can someone give me a quick explination for how the DB frame data is layed out? I'm assuming the second hit hitting on 27 is the FASTEST it can be done right? And that frame 15 is the EARLIEST it can cancel into the next hit right? Please confirm/deconfirm.

I need this information because I'm reworking Marth's DB for our "Roy project" in brawl+, so I need to know which frames I'd need to slow down/speed up to change when he can continue to the next hit, as well as which frames I'd change to change the ending lag on the individual hits (without continuing). Please and thank you.

 

[Yoshi Kirishima]

Oh i see.

I'm not sure about that .

Roy in Brawl+? Awsome :D Good luck!

 

[JUDGE]

good job steel

 

[Shaya]

Dark Sonic,

I'm completely redoing Marth's frame data from scratch. I'll be fixing up and adding "Jab windows" to db's data.

Hit Lag (or as often referred to as Shield Hit Lag), is the count of how many frames after the move hits (so if it hits frame 4, frame 5 = 1, not 2). Shield Stun is counting from the same point until the opponent can 'shield grab' (and at the same time they could jump, roll, spot dodge, up b, up smash, etc).

There's no need to over complicate things with shieldhitstun(Marth) or shieldhitstun(opponent). Hit lag and shieldstun are constant and don't really need any other specifics to be counted.

As I'm sure you already know this,

The only time this doesn't apply is if the hitbox is seen as a projectile/windhitbox (and wind hit boxes have their own bugs, i.e. counter jacket), where the hitlag won't apply to the hitter only the receiver.

 

[BarDulL]

Dark Sonic,

I'm completely redoing Marth's frame data from scratch. I'll be fixing up and adding "Jab windows" to db's data.

hurry up already

 

[Steel]

srsly

for someone with debug mode you sure take weeks months for this stuff

 

[Shaya]

Well I'm sorry, I have 500 years of Uni work to worry about.

Plus I'm doing something too awesome with zeh data.

Examples:

Up Taunt:
End: 84

Side Taunt:
End: 104

Down Taunt
End: 79

Oh but seriously...

Jab 1
Hit: 4-7
End: 27
Hit Lag: 3
Hit Lag (tip): 8
Shield Stun: 6
Shield Stun (tip): 8
Shield Advantage: -20
Shield Advantage (tip): -23

Jab Window: 19-26 (5 frame start up) / 27~29 (4 frame start up)
Hit Stun End: 225
Hit Stun End (tip): 83
Hit Stun Neutral: 141
Hit Stun Neutral (tip): 38

Hit stun values are based on Mario.
'End' is where the max hit stun (30 frames) will be reached [damage percentage], and further damage will allow tumbling cancelling out hitstun at frame 13 with an airdodge/shield, or 25 with an attack.
'Neutral' is the damage required where the hitstun will have you at a 0 frame disadvantage.

I'm doing this for all moves except those who start out with tumble, and some who barely have time before going into tumble (utilt starts at around 26 frames of hitstun, and will start tumbling from 5%, i.e. useless for combaws).

Oh, and I guess I'll show dtilt too:

Down Tilt
Hit: 7-9
IASA: 21
Hit Lag: 5
Hit Lag (tip): 10
Shield Stun: 10
Shield Stun (tip) 10
Shield Advantage: -9
Shield Advantage (tip): -14

Hit Stun End: 89
Hit Stun End (tip): 52
Hit Stun Neutral: -41
Hit Stun Neutral (tip): -69

As the neutral is 'below 0', it insinuates that it's epicz for combawz from 0.

Oh, and DB1 to grab, eventually ALMOST becomes a real combo (at like 400%)

 

[Dark Sonic]

There's no need to over complicate things with shieldhitstun(Marth) or shieldhitstun(opponent). Hit lag and shieldstun are constant and don't really need any other specifics to be counted.

Actually, they are NOT constant. It is CLEARLY indicated that the opponent was out of hitlag before Marth was (the opponent slid, while Marth was STILL frozen). Magus noticed this on Shiek's f-tilt a long time ago, and I just posted it on Marth's dash attack (you can check my data yourself if you like).

So you DO need to list them, as they are actually different.

As I'm sure you already know this,

The only time this doesn't apply is if the hitbox is seen as a projectile/windhitbox (and wind hit boxes have their own bugs, i.e. counter jacket), where the hitlag won't apply to the hitter only the receiver.

This is false, as my data clearly indicates. I used the frame speed code and simply counted the number of frames passed before the opponent started to slide after blocking the attack (and noticed that Marth was still frozen, so I took note of the opponent and continued counting for Marth). For shieldstun itself, I counted until the opponent was able to drop their shield (the shield itself disappears on frame 1, despite them not being able to move). I tested and retested and every time I got these results. Tippered shieldstun was higher, tippered shield hitlag was higher (while being disproportionately higher for Marth), and the hitlag for Marth was not the same as the shieldhitlag for the opponent.

Hitlag always being the same value as shieldhitlag for the opposing player was debunked months ago.

To repeat my method so you can get repeatable results yourself, this is how I tested each stat. (I however included the frame that the move hit) I used the frame advance code to slow the game down frame by frame.

Shieldstun-counted from when the move hit until the opponent dropped their shield (the shield disappears on frame 1. A very nice visual cue).

Shieldhitlag {noted as shieldhitlag(opponent)}-counted from when the move hit until the opponent started sliding from their position (they're frozen in place during shieldhitlag, and the shield coincidentally doesn't change color either. Two very easy visual cues).

Hitlag{noted as shieldhitlag(Marth)}-counted from when the move hit until Marth continued movement.

Startup lag (counting until the move hits)
Total lag (counting until you can shield. I whiffed the move for this)
Duration (having the opponent come out of spotdodge invincibility on the desired frame for testing, bumping up 1 frame at a time until it no longer hits).

And to repeat one final time, shieldhitlag and hitlag are not necessarily the same. Verify it yourself if you like. Magus stated this a LONG time ago when discovering that Sheik suffers less than HALF as much hitlag as a shielding opponent when she hits a shield with her f-tilt. If you don't trust my results then test it yourself, but don't just say that my results are wrong because it doesn't support the old theory (which has been debunked). Also, untipped hitlags (both of them) are different from tipped hitlags.

 

[Steel]

dark sonic >>>> shaya

 

[Dark Sonic]

Also Shaya I've been meaning to ask....how'd you get a negative number for hitstun? Are you trying to do some kind of formula based off of the BKB and KBG of the attacks? (cause we can give you those numbers).

I don't remember the formula myself (Leaf's the one who's good with that. He even figured out how character specific division constants are factored into hitstun).

Btw note that each character can suffer different amounts of hitstun even when applied with the same launch speed. As in, some characters LITTERALLY suffer less hitstun than others, even when the forces applied to both of them are constant. Any hitstun values you derive will be GENERAL hitstun values, as they will vary slightly from character to character.

 

[Shaya]

I'm not recording hitstun values, I'm using a simple formula [that for most characters seems to work] to calculate when a move becomes safe to hit with (percent wise) (i.e. hit - end + hitstun = 0), and it's based on Mario. Some characters are weird (I've noticed) but the general thing has shown that most character's hit stuns will never go above 30, and it reaches this cap at 2550 launch speed (from what I've seen so far).

I was never aware of the differed hit lag values, but I'm still confused about their relevance. The main point of hitlag is to calculate shield advantage (at this time, at least).
So if Marth suffers 6 frames of hitlag, and the opponent suffers 11 frames of shieldstun, the relevance of recording shieldhitlag opponent, when the two run concurrently seems.. redundant. I do know of it from when testing frame data, noting that characters will only start to 'move' naturally after a certain point... I suppose it could be an indicator of how 'well' it can keep you safe if you hit their shield...

However, the move having different hit lag values for whether it hits a shield or not is of much higher importance, but I've not noticed such discrepencies doing Marth's frame data as of yet. However, the shield values are in general of much higher importance..

Hence why I say it's of not much importance.

 

[Dark Sonic]

Come on man...you know what hitlag IS right? Shieldhitlag functions the same as hitlag (even though the values are different). Here's an extremely easy way to see the significance.

Use the timer item and have Ganondorf do a shorthop dair on the slowed opponent (making sure to buffer the dair of course). Now, you should notice that the opponent is still in hitlag while you're not in lag. BINGO, you've just gained a large advantage because of a large hitlag difference

Take that concept and apply it on a smaller scale and you'll see my point.

I was never aware of the differed hit lag values, but I'm still confused about their relevance. The main point of hitlag is to calculate shield advantage (at this time, at least).
So if Marth suffers 6 frames of hitlag, and the opponent suffers 11 frames of shieldstun, the relevance of recording shieldhitlag opponent, when the two run concurrently seems.. redundant.

SHIELDSTUN AND HITLAG DO NOT HAPPEN AT THE SAME TIME! Hitlag (and shieldHITLAG) happen BEFORE shieldstun. They are two seperate events. Ergo, if you have an advantage in hitlag, THAT WILL ADD TO YOUR TOTAL ADVANTAGE. It needs to be added into your calculation.

However, the move having different hit lag values for whether it hits a shield or not is of much higher importance, but I've not noticed such discrepencies doing Marth's frame data as of yet. However, the shield values are in general of much higher importance..

Hence why I say it's of not much importance.

Shieldhitlag and shieldstun are different.

This is how it works.

Marth's dash attack hits shield (not tipped). Marth suffers 6 frames of hitlag while the opponent suffers 8 frames of shieldhitlag, giving you AN ADDITIONAL 2 FRAMES towards your advantage (because Marth is actually out of hitlag 2 frames before the opponent, and they haven't even gone through shieldstun yet). Opponent suffers 11 frames of shieldstun.

Using YOUR method of calculating advantage, you would've incorrectly calculated that Marth's advantage on a shielded nontipped dash attack is -25. However, it is ACTUALLY -23 because Marth has 2 extra frames where his opponent is still in hitlag while he is not (BEFORE shieldstun).

Now, looking at the tipped dash attack you've got

Marth's dash attack hits shield (tipped). Marth suffers 12 frames of hitlag while the opponent only suffers 9, CUTTING INTO YOUR ADVANTAGE. Opponent suffers 14 frames of shieldstun.

You'd have calculated the advantage to be -22, when the advantage is ACTUALLY -25 (you suffer an extra 3 frames of hitlag while your opponent starts cutting into their shieldstun), making it WORSE than an untipped dash attack.


THAT is why you need to take note of the hitlag differences. The extra hitlag by either side affects Marth's advantage when the move is blocked. If Marth suffers less hitlag then the opponent, then he is cutting into his ending lag before the opponent moves to shieldstun and gains advantage. But if the opponent suffers less hitlag than Marth, then they're cutting into their shieldstun before Marth even goes into his ending lag. Every one of your advantage calculations would be off by a couple frames. Though that may not matter with things like dash attack, I'd think it would matter with oh...d-tilt, fair, nair, ect, ect.

 

[Shaya]

Hmm.

So what you're saying is,

when I, in pause code mode do the following:

Have one character shield
have another dash attack into shield
then time (after the hit) how many frames it takes for
a) The animation for Marth's dash attack to CONTINUE
b) How long until a SHIELD GRAB can start

You're now saying this data, which would simply show how many frames after a hit on shield will it take for Marth to be able to do another action is.... wrong?

While it seems our data differs slightly (probably the way we count, I have DA's hit lag at 5, shieldstun at 10), you saying that Marth's dash attack is only -22 or -25... is well, nothing close to what I have.
I have advantage for untippered dash attack at -31. In other words how many frames after a hit on shield can have a shield grab START to when Marth can shield again.

Hence there is a much LARGER confusion here, beyond just what the definition of hitlag is.

I can however understand the difference in hitlag being important if it actually does hit (i.e. it constitutes as something AKIN to hitstun), but unless there's something I'm missing for calculating shield hit advantage...

Perhaps you wouldn't mind link me to a topic stating what you have told me? Because I am pretty sure that Marth's dash attack is not only -21/22 on shield...

 

[Magus420]

Why do people write Shield Hitlag+Shield Stun as "Shield Stun"? It's redundant and lends itself to confusion and errors (such as DS including Shield Hitlag twice by mistake when calculating the advantage). The opponent's Shield Hitlag would already be listed on its own. The shieldstun on the untipped looks like it should be 3, and not 11.


Shield Hitlag Differential = Shield Hitlag (defender) - Shield Hitlag (attacker)
Advantage = Shield Hitlag Differential + Shield Stun - Remaining Frames of the Move

Going off DS's numbers it should really be listed more like this imo:

Dash Attack
Total: 49
Hit: 13-16

Un-Tipped
Hitlag: (6?)
Shield Hitlag: 6 (Marth); 8 (opponent); SHD: +2
Shield Stun: 3
Advantage: -31 to -28

Tipped
Hitlag: (12?)
Shield Hitlag: 12 (Marth); 9 (opponent); SHD: -3
Shield Stun: 5
Advantage: -34 to -31

 

[Dark Sonic]

b) How long until a SHIELD GRAB can start

No, no, no. You need to count until the shielding character starts sliding (to indicate that they're out of shieldHITLAG). However, this character has just now entered shieldstun and still can't move, so you should count how many frames the opponent has until they can DROP THEIR SHIELD (the shield disappears on frame 1.) You can alternatively count until they can grab, but you have to make sure to subtract the amount of frames it takes...to actually grab (since in order to grab you have to be OUT OF SHIELDSTUN in the first place).

edit: Actually, since the shield disappears when they start their grab anyway, that should work (same with spotdodges and rolls).

You're now saying this data, which would simply show how many frames after a hit on shield will it take for Marth to be able to do another action is.... wrong?

It would be correct for MARTH, however, it would not be correct for the opponent. Advantage is the difference between when Marth can move and when the opponent can correct? Well here's an extra 2-3 frames where one is moving and the other is not.

While it seems our data differs slightly (probably the way we count, I have DA's hit lag at 5, shieldstun at 10), you saying that Marth's dash attack is only -22 or -25... is well, nothing close to what I have.

I miscounted. I counted hitlag into shieldstun by mistake (in addition to adding 1 to everything for some reason )

I have advantage for untippered dash attack at -31. In other words how many frames after a hit on shield can have a shield grab START to when Marth can shield again.

I miscounted. However, the principle remains the same.

The hitlag differential is added to Marth's advantage because it essentiall acts as additional lag to one of the characters.

My miscounting was me including the hitlag in shieldstun, therefore counting it twice...my bad.

 

[uhmuzing]

I'm sorry, but this is confusing me

ADVANTAGE FORMULA: SHIELDSTUN + HIT - SHIELDHITLAG - END (or IASA)

 

[Dark Sonic]

I'm sorry, but this is confusing me

ADVANTAGE FORMULA: SHIELDSTUN + SHIELDHITLAG - HITLAG - END (or IASA)

Fixed.

Hitlag=time Marth is frozen. Shieldhitlag=time enemy is frozen. Shieldstun=time enemy is unable to take action (after shieldhitlag)

 

[Shaya]

Okay, so whilst my use of 'shield stun' is incorrect; my calculations of advantage are... not?

Edit: Ooops I'm an idiot, the 'hit length' didn't phase me.

Final question, when the move actually hits the opponent, am I to presume that the hit lag (marth)/hit lag(opponent) aren't actually the same?

 

[Dark Sonic]

Final question, when the move actually hits the opponent, am I to presume that the hit lag (marth)/hit lag(opponent) aren't actually the same?

When they actually hit the opponent they are the same, however when you hit a shield they are not.

 

[Shaya]

Now to find a great term for '(Shield hit lag + shield stun)'

Shield Stun Lag!
SHIELD HIT STUN ohdear...

Oh well.

Actually I may end up going about finding the individual shield hit lags, if and only if that something like 'shield hit lag' remained constant with decayed moves whilst shield stun did not... (or something similar).

 

[Pikabunz]

Is it me or is the advantage formula wrong? I think Magus has the right one.

 

[Steel]

our formulas are the same

but you could say his is more clear though, but it doesn't matter. everyones frame data is getting redone sometime in the future by the smash lab

 

[Sake-Hato]

is it me or dose Mk's Fair Beats marths? o.O