Data Kadano's perfect Marth class -- advanced frame data application

[net1234]

Wow dancing blade is really bad

 

[Kadano]

Wow dancing blade is really bad

It’s somewhat good if it connects, but on shields, it’s pretty bad, yeah. Unless your opponent doesn’t know his options, but in that case, about anything works.

 

[net1234]

This is probably a stupid question but I can't find a falco cg guide anywhere

 

[Mahie]

Thanks for the Falcon case, it's mostly vs him/fox/sheik that I use it anyway.

I guess not connecting with DB is even worse than I thought.

It's good knowing that they can't roll behind, though.

 

[hectohertz]

@ Kadano i was playing around with sheiks aerial interrupts, i've found you can use them to land on a platform with her by doing a uair exactly 7/9 (i forget which) frames after jumping (on bf + DL).

is there anyway to use this from the ledge, like samus can?

 

[FrogMarth]

hey kadano or anyone.but hopefully kadano. im a new marth main and ive been working on comboing. i know you posted the hitbox of the fair but i was wondering if it would be possible to show the different angles that characters go flying based on what part of the hit bubble you hit them with like you did with the up tilt and dancing blade. im asking because ive noticed that when im attempting the ken combo tthe character goes straight up when i tip the fair and sometimes it goes more diagonally up... or is that just d.i.? id also appreciate the same with dair and nair if its possible. looking forward to your new melee mechanics vids especially the movement one

 

[Bones0]

hey kadano or anyone.but hopefully kadano. im a new marth main and ive been working on comboing. i know you posted the hitbox of the fair but i was wondering if it would be possible to show the different angles that characters go flying based on what part of the hit bubble you hit them with like you did with the up tilt and dancing blade. im asking because ive noticed that when im attempting the ken combo tthe character goes straight up when i tip the fair and sometimes it goes more diagonally up... or is that just d.i.? id also appreciate the same with dair and nair if its possible. looking forward to your new melee mechanics vids especially the movement one

http://www.mediafire.com/view/kwbkgsdok1dekd7/SSBM_Hitboxes_(NTSC_1.0)_&_Knockback_1.5k.xlsx

Go to the first sheet, and scroll down to A2266. Assuming I'm reading it correctly, fair has 4 different hitboxes. The tip sends at 67 degrees while the rest send at 361. Not sure exactly what the 361 means since that seems the same as a 1 degree angle...

 

[Kadano]

@ Kadano i was playing around with sheiks aerial interrupts, i've found you can use them to land on a platform with her by doing a uair exactly 7/9 (i forget which) frames after jumping (on bf + DL).

is there anyway to use this from the ledge, like samus can?

Nice find! Yeah, it’s +8. Fast fall for 4 frames, jump for 11, then uair. I’ll add it to the chart.

hey kadano or anyone.but hopefully kadano. im a new marth main and ive been working on comboing. i know you posted the hitbox of the fair but i was wondering if it would be possible to show the different angles that characters go flying based on what part of the hit bubble you hit them with like you did with the up tilt and dancing blade. im asking because ive noticed that when im attempting the ken combo tthe character goes straight up when i tip the fair and sometimes it goes more diagonally up... or is that just d.i.? id also appreciate the same with dair and nair if its possible. looking forward to your new melee mechanics vids especially the movement one

http://www.ssbwiki.com/Marth_(SSBM)/Forward_aerial
Click on the top image to the right for the most visual explanation.

The tip sends at 67 degrees while the rest send at 361. Not sure exactly what the 361 means since that seems the same as a 1 degree angle...

http://www.ssbwiki.com/Sakurai_angle

 

[FrogMarth]

soo there wouldnt be a difference of angle between tipping with the top of fair and tipping with the bottom?

 

[Kadano]

soo there wouldnt be a difference of angle between tipping with the top of fair and tipping with the bottom?

Yes.

 

[hectohertz]

Nice find! Yeah, it’s +8. Fast fall for 4 frames, jump for 11, then uair. I’ll add it to the chart.

http://www.ssbwiki.com/Marth_(SSBM)/Forward_aerial
Click on the top image to the right for the most visual explanation.


http://www.ssbwiki.com/Sakurai_angle

<3 i did something useful! any chance i could get a gif of it?

 

[SpiderMad]

That's really kind of you two, but the problem was caused solely by my own dumbness / clumsiness, so I'll pay for it by myself. I did apply for the YouTube donation button though; if it doesn't work out, I might do a Paypal comission thing if there's interest.

@ hectohertz : Resolution is not the problem. I'd rather do your requests like I've been doing my videos, if that's fine with you.

@ZoSo: I started looking into SDI on Fox's uair and it seems single SDI is never sufficient if the Fox has good execution. At least at 0%. :/

What did you do to have caused a broken modem?

 

[Bones0]

What did you do to have caused a broken modem?

He uploaded too many HQ gifs. A real modern day Smash tragedy.

 

[Kadano]

<3 i did something useful! any chance i could get a gif of it?

Of course. I’m halfway done, just need to finish the input display. I gotta leave now, but I’ll deliver in the evening.

What did you do to have caused a broken modem?

It’s a modem/router that gets its data from mobile radio signals. Thus, I kept it on a window board so the data only had to move through glass and not concrete. When I tested all possible locations in my house, this one got the fastest connection. (20 Mbit/s down, 4 Mbit/s up are the top rates possible there.)

The handle of that window was broken; whenever I pulled on it, it would come off. So I kept it at the opened position and just pushed the window into its frame when I wanted to have it closed. I should mention that it was very hot at that time (27°C, no idea how much that is in medieval units) so I had all my windows open during the night.

I went on a vacation for one week. During that week, it was very rainy and stormy. When I came home, the window had been blown open and it had rained on the modem, which was still powered on as my flatmate was still at home for part of that week. On the outside of that window, the roof is very protruding, so the raindrops had to fall with about 45° in to reach the modem.

It’s completely my own fault for not shutting the window completely. But when I packed my stuff, I was so busy getting all of my Melee testing hardware together that I didn’t think about it. (During the one week vacation, I didn’t have internet and created the video “Tumble and jump buffering”.)

 

[MookieRah]

Just going to point out that in Europe, a lot of housing and apartments have amazing insulation, and as such it's not as common of a thing to have air conditioning (hence opening windows in the first place).

 

[hectohertz]

On the outside of that window, the roof is very protruding, so the raindrops had to fall with about 45° in to reach the modem.

this is the most kadano explanation ever. he's got frame data for the rain <3 <3 <3

 

[Kadano]

he's got frame data for the rain <3 <3 <3

Well … here’s some proper frame data:

AttackS3S 8 is the last frame of intangibility, by the way. I think I’ll either go for hybrid hitbubble visibility in the future or add an overlay that indicates the current body state.
(Wow, gfycats embedded here with donotresize checked look just as fine as I always wished for them to do. Writing posts here is gonna be a lot more fun now. By viewing it at their site you can see it in native 1440×1080p60 resolution btw.)

 

[1MachGO]

Hey Kadano, is fthrow>dash pivot fsmash a legit follow up/frame trap on the following characters? And if so, what percents?:

Pikachu
Ice Climbers
Mario/Doc
Luigi

 

[hectohertz]

Well … here’s some proper frame data:

AttackS3S 8 is the last frame of intangibility, by the way. I think I’ll either go for hybrid hitbubble visibility in the future or add an overlay that indicates the current body state.
(Wow, gfycats embedded here with donotresize checked look just as fine as I always wished for them to do. Writing posts here is gonna be a lot more fun now. By viewing it at their site you can see it in native 1440×1080p60 resolution btw.)

amazing stuff as always. btw apparently pika has an aerial interrupt as well with fair and uair.

 

[Kadano]

amazing stuff as always. btw apparently pika has an aerial interrupt as well with fair and uair.

I can’t find it. Both attacks have little downward ECB shift to compensate for Pikachu’s high vertical acceleration, so I doubt these could produce aerial interrupts with substantial gaint.

 

[Kadano]

Hey Kadano, is fthrow>dash pivot fsmash a legit follow up/frame trap on the following characters? And if so, what percents?:

Pikachu
Ice Climbers
Mario/Doc
Luigi

Testing this is tedious. For it to be precise, I need to do at least 16 DI samples for each percent, throw and follow-up combination. I quickly went over Pikachu, and some of the follow-ups on him that are guaranteed on specific DI include tipper fsmash, pivot tipper fsmash, tipper dsmash, dash attack, dash sh fair. So that’s at least 6. Multiply that by 16 for the minimum DI samples, then by two for both fthrow and dthrow. Thus, for every percent, I need to test 192 different option sets for every single percent to get an exact output graph like I did for Jigglypuff (only did 90% there and that took a few days, so what you are asking for is much more).

(Also, yes, I do understand you only asked for fthrow → pivot fsmash, but isolating this option is futile as half of all DI angles escape it every single time. Fthrow → pivot fsmash is always beaten by good DI if it is the only threat you bring in. Only with threats that counter the DI that escapes it does it become a useful tool.)

The motivations for doing so aren’t strong enough, so I won’t. I can tell you what I’m guessing from testing the options that seemed most likely to me:
1. fthrow fsmash works at low to mid percent against Pikachu if he DIs up.
2. dtrow turn dsmash works at low to mid percent if he DIs down.
3. fthrow pivot fsmash works at mid percent if he DIs slightly in.
4. fthrow dash attack works at mid percent if he does not DI.
5. 287° DI escapes every follow up at all percents I tested it at.

 

[Bones0]

Kadano, have you ever explored the possibility of DDing into a pivot dsmash so that you can do it off of fthrows? Like fthrow right, dash right, dash left (very short distance), pivot right, dsmash. DI down and away is pretty standard for getting grabbed by Marth, so if this could somehow be used to KO floaties, I think it'd be worth the difficulty (it's actually not too bad; I actually used to think this was the only way to fsmash the same direction you dashed so I practiced it for a while lol).

 

[hectohertz]

I can’t find it. Both attacks have little downward ECB shift to compensate for Pikachu’s high vertical acceleration, so I doubt these could produce aerial interrupts with substantial gaint.

hmm interesting. i'll play around in debug mode tn, will report if i find anything. also apparently samus has a zero-impact-land / no-impact-land / lagless land (we need a standard term for this tech). im gonna test it tn

 

[Kadano]

Kadano, have you ever explored the possibility of DDing into a pivot dsmash so that you can do it off of fthrows? Like fthrow right, dash right, dash left (very short distance), pivot right, dsmash. DI down and away is pretty standard for getting grabbed by Marth, so if this could somehow be used to KO floaties, I think it'd be worth the difficulty (it's actually not too bad; I actually used to think this was the only way to fsmash the same direction you dashed so I practiced it for a while lol).

Yeah, I did. I never found a throw setup for it, though.

 

[Bones0]

hmm interesting. i'll play around in debug mode tn, will report if i find anything. also apparently samus has a zero-impact-land / no-impact-land / lagless land (we need a standard term for this tech). im gonna test it tn

I have always called it "aerial landing", but that's pretty vague and boring so I'm not opposed to a new term. No impact landing is when you DJ at the perfect height to land on a platform so that you have no lag, so we can't use that.

 

[hectohertz]

I have always called it "aerial landing", but that's pretty vague and boring so I'm not opposed to a new term. No impact landing is when you DJ at the perfect height to land on a platform so that you have no lag, so we can't use that.

that's what im talking about. you can NIL onto the stage with many characters, from testing with sheik she cant retain any invincibility. apparently samus can.

 

[Kadano]

hmm interesting. i'll play around in debug mode tn, will report if i find anything. also apparently samus has a zero-impact-land / no-impact-land / lagless land (we need a standard term for this tech).

I’m fine with abandoning the “lagless land” term. To be precise, it is not even lagless – there’s always one single frame of landing lag (Wait 0 ). I think “no-impact land” is fine. Should we use “impact land” only for the 4-6 frames of standard landing lag or also for AttackAirX landing lag (that can be L-canceled)?

you can NIL onto the stage with many characters, from testing with sheik she cant retain any invincibility. apparently samus can.

What is it with you and the word “apparently”? I think every single time I witnessed you using that word, the associated statement was wrong, haha.

I tried to get Samus to perform a lagless land (with gaint>0) from the ledge, but I’m pretty sure it’s impossible. I even tested on different stages, but in none of them the ECB acts favorably.

 

[hectohertz]

I’m fine with abandoning the “lagless land” term. To be precise, it is not even lagless – there’s always one single frame of landing lag (Wait 0 ). I think “no-impact land” is fine. Should we use “impact land” only for the 4-6 frames of standard landing lag or also for AttackAirX landing lag (that can be L-canceled)?


What is it with you and the word “apparently”? I think every single time I witnessed you using that word, the associated statement was wrong, haha.

I tried to get Samus to perform a lagless land (with gaint>0) from the ledge, but I’m pretty sure it’s impossible. I even tested on different stages, but in none of them the ECB acts favorably.

i like no-impact-land (NIL) as well. i like impact land for only the standard landing lag.

hahahaha damn you kadano, reading all my language! i use 'apparently' when someone tells me something that I have not verified. which (as you pointed out) tends to be wrong. people are scrubs.

yeah i tested samus' NIL for a while last night, i also can confirm that I could not get any gaint on it.

i might play with pikachu tonight, and see if i can find anything. i suspect you are better than this than me, but you never know

 

[hectohertz]

I’m fine with abandoning the “lagless land” term. To be precise, it is not even lagless – there’s always one single frame of landing lag (Wait 0 ). I think “no-impact land” is fine. Should we use “impact land” only for the 4-6 frames of standard landing lag or also for AttackAirX landing lag (that can be L-canceled)?


What is it with you and the word “apparently”? I think every single time I witnessed you using that word, the associated statement was wrong, haha.

I tried to get Samus to perform a lagless land (with gaint>0) from the ledge, but I’m pretty sure it’s impossible. I even tested on different stages, but in none of them the ECB acts favorably.

okay i found an aerial interrupt with gaint with pikachu (only tested on fd so far):
FF for 6 frames,
jump for 12 frames
uair

edit: i think its ECB dependent. i leave the details to you @ Kadano

 

[1MachGO]

Hey Kadano, sorry if this is off topic, but I have some questions concerning the mechanics which determine the lengths of initial dashes/dash dances. Is it all determined by initial dash speed accelerating (or decelerating in some cases) to run speed? Or does traction play a role? Furthermore, does every character have the same # of frames for their initial dash?

 

[Kadano]

Hey Kadano, sorry if this is off topic, but I have some questions concerning the mechanics which determine the lengths of initial dashes/dash dances. Is it all determined by initial dash speed accelerating (or decelerating in some cases) to run speed? Or does traction play a role? Furthermore, does every character have the same # of frames for their initial dash?

Afaik there’s only dash init and run speed (see Toomai’s hitbox collection; it has a char values tab). For dash frame data, see this excellent post by Stratocaster: http://smashboards.com/threads/2014...nowledge-updated-1-2-14.339520/#post-16153183

 

[Kadano]

The video on teching is finished!

https://www.youtube.com/watch?v=ZkdPVUlrSOo

 

[MookieRah]

Something I wanted to note about the video. When you want to do a wall jump from a wall tech you can just hold up on the control stick and get it every time. I assume this works holding X or Y as well. Technically I could just be hitting it within the window of time, but I never seem to pull it off with consistency if I don't hold up.

Also, about your note at 13:57 regarding the properties of Falcon's knee, does this apply to other moves as well? By that I don't mean that other moves have the property to lift you off the ground, but moves that send you at higher angles that otherwise would put your character in the air? If so, that could explain why I'd always accidentally tech things throughout my M2 career (to the point people thought I was doing that **** on purpose xD).

EDIT: For instance, I'm pretty sure I've been able to tech to the ground from Marth tippers just by inputing my shffl with M2.

In any case, this is definitely illuminating, and I'm really hoping that ECB thing applies to more than just Falcon's knee for floor techs. Amazing video as always.

 

[hectohertz]

The video on teching is finished!

https://www.youtube.com/watch?v=ZkdPVUlrSOo

awesome video as always!

a question:

if you get a double-stick tech, and have enough momentum to slide offstage (and are facing the wrong way to trigger ottotto), does fastfalling work as an input to cancel momentum during the one frame window where you can input an attack? i've had success double stick teching then FFing and grabbing the ledge, but im unsure of the exact mechanics of why this works (perhaps i just had a relatively small amount of momentum).

 

[Kadano]

Something I wanted to note about the video. When you want to do a wall jump from a wall tech you can just hold up on the control stick and get it every time. I assume this works holding X or Y as well. Technically I could just be hitting it within the window of time, but I never seem to pull it off with consistency if I don't hold up.

No, held X or Y inputs don’t trigger PWJ unless you start them late enough to trigger the 20 frames jump buffer.
During the first 4 frames of PassiveWall, control stick position and fresh X/Y inputs are polled. Held X/Y inputs are not polled. (I’m not sure if poll is the right term here, but I don’t know a better one.)
I don’t recommend relying on holding the control stick upwards as doing so makes it harder to get the maximum horizontal aerial control towards the stage by holding in that direction. Pressing X/Y at the same frame (more or less) as the tech input seems the easiest and most efficient to me.

Also, about your note at 13:57 regarding the properties of Falcon's knee, does this apply to other moves as well? By that I don't mean that other moves have the property to lift you off the ground, but moves that send you at higher angles that otherwise would put your character in the air? If so, that could explain why I'd always accidentally tech things throughout my M2 career (to the point people thought I was doing that **** on purpose xD).

EDIT: For instance, I'm pretty sure I've been able to tech to the ground from Marth tippers just by inputing my shffl with M2.

I don’t understand your question, but I’ll simply everything there is to this, hopefully this will answer it.
For DSDI, there are only three relevant hit attributtes: amount of knockback, base trajectory and amount of hitlag.
The higher the knockback, the more distance has to be compensated with TDI and ASDI at a given trajectory.
The higher the base trajectory, the greater the upwards part of the knockback vector. So basically we can approximate: [amount of TDI and ASDI needed]=[knockback]*sin[base trajectory].

The amount of hitlag only matters in that every character that is not DK, G&W, Jigglypuff, Kirby or Mewtwo needs to SDI back to the ground near-perfectly if hitlag is 10 or greater. As hitlag is calculated from hit damage, charging and staling can make a difference here. For example, Marth’s fresh tipper fsmash causes 9 frames of hitlag uncharged, but 10 frames of hitlag if slightly charged. (I think it takes about 20 frames of charging to get to 10 frames of hitlag, but that’s only a guess.)
A completely stale knee only causes 9 frames of hitlag, so SDI would not be necessary for any characters.

if you get a double-stick tech, and have enough momentum to slide offstage (and are facing the wrong way to trigger ottotto), does fastfalling work as an input to cancel momentum during the one frame window where you can input an attack? i've had success double stick teching then FFing and grabbing the ledge, but im unsure of the exact mechanics of why this works (perhaps i just had a relatively small amount of momentum).

That only works on low momentum.

 

[Bones0]

Amazing video, and in just a day you've already got 1,000 views! :D

A few questions that can hopefully be answered pretty easily:

1. Are attacks (just A moves?) and taunts the only things you can use on the frame before slideoffs? I thought maybe I could use tap jump's 3-frame buffer. If not, the best course of action might be to just piano-key a bunch of different inputs 1 frame after another (A for jab, B for shield breaker, C-stick for fsmash, Z for grab). Ooh, actually, is there any chance you could just spin the C-stick to guarantee at least a smash comes out?

2. To get the 305 degree SDI on Falcon's knee, can I quarter circle (starting from down) so that I do multiple SDI inputs and the game does the first one that isn't forbidden? Also, why are those select few chars different from the rest of the cast? I thought maybe it was due to low weight, but then DK was in there so idk.

3. When you did the MSDI for Marth's fsmash, the first SDI input was slightly above the pink arrow, then you went to slightly below the blue arrow. Isn't the easiest way to get 3 SDIs to go from up-left to left to down-left in one motion?

 

[Kadano]

1. Are attacks (just A moves?) and taunts the only things you can use on the frame before slideoffs? I thought maybe I could use tap jump's 3-frame buffer. If not, the best course of action might be to just piano-key a bunch of different inputs 1 frame after another (A for jab, B for shield breaker, C-stick for fsmash, Z for grab). Ooh, actually, is there any chance you could just spin the C-stick to guarantee at least a smash comes out?

No B moves. Jumping doesn’t work either – remember how Fox and Falco can do their shinehog? It consists of sliding off the stage during kneebend, so from that we know that kneebend / jumping doesn’t prevent sliding off.
C-stick spinning doesn’t help much as it takes quite far input differences to trigger an additional fresh smash input.
I think the optimal thing to do would be to press up on the d-pad, A, Z and C-stick left (with thenar grip) at about the same time.

2. To get the 305 degree SDI on Falcon's knee, can I quarter circle (starting from down) so that I do multiple SDI inputs and the game does the first one that isn't forbidden? Also, why are those select few chars different from the rest of the cast? I thought maybe it was due to low weight, but then DK was in there so idk.

No, quarter circle DI doesn’t help because 305° is not one of the key angles. Although the name “forbidden SDI” invokes the idea of the game preventing the SDI itself, it’s actually a bit different. The SDI is still applied, but immediately corrected back to your previous position.

Mewtwo, DK, Jigglypuff etc. have the same downward ECB extension in their knockback/hitstun state as in their grounded state, so their feet don’t lift off the ground, so to speak.

3. When you did the MSDI for Marth's fsmash, the first SDI input was slightly above the pink arrow, then you went to slightly below the blue arrow. Isn't the easiest way to get 3 SDIs to go from up-left to left to down-left in one motion?

No, doing so would only yield two SDI outputs. And that while moving a whole 90° — the way I did it took only ~60°. It was not quite optimal, though – 180°→197°→163° would have been even better both in total SDI distance and in total angle movement span.

 

[tm]

Really great video. These are getting better and better

I left some questions in the comments, I'll repost them here and copy your answers to whichever place you leave out.


"I believe it is often plausible to fastfall to ledge when you are about to slide off backwards, and that this may be easier than momentum stalling (easier to time / larger window?)"
-- you did already address this to Hectohertz, but I have a hard time believing that this isn't practical even at high knockback, at least for fastfallers. Would it be too much to ask for you to look into this a bit more and come back with some numbers? Not a problem if it's too much work.


"Also, what is the joystick UP value threshold for starting the walljump during a wall tech? If the joystick is being help up and inwards, what degree angle would that be?"

 

[1MachGO]

Kadano, if ASDI and TDI can be used to force yourself into the ground, can it also be used to force yourself into the air? Is it humanly possible to force yourself airborne vs. Fox's shine, land cancel hit stun, and get out of follow ups?

 

[tm]

Kadano, if ASDI and TDI can be used to force yourself into the ground, can it also be used to force yourself into the air? Is it humanly possible to force yourself airborne vs. Fox's shine, land cancel hit stun, and get out of follow ups?

As stated in the video and the DI thread, smash DI cannot put you on the ground from midair nor can it put you in the air from the ground. So no