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

[tm]

Yeah that's true, but you'll at least keep them honest. Also a lot of characters won't beat marth to ledge because it's very hard (if even possible) to WD to ledge from that position, so they will have to short hop fastfall (obviously fastfallers can do this easily, but you're usually dead from that edgeguard position vs them anyway). It's definitely not foolproof but it will probably work at least once on any non-fastfaller who hasn't practiced punishing it

 

[Bones0]

You can WD to the ledge even if you roll all the way next to it, but you have to be frame perfect or you die. But that's just Melee in a nutshell, I guess.

 

[Omegaspike_]

This is really good, and it's cool to have somewhere to go for an extremely in depth guide c:

 

[Tityboi]

alright............ I swear to god I just grabbed a fox during the bounce of missed tech...... Is this possible, or am I imagining it?

 

[SpiderMad]

@ Kadano @ Bones0
Do you guys feel like Melee made the right decision to restrict Specials (and Air dodge) from being able to be used during IASA frames (when in the air)? Not having this restriction would allow SH Bair WL and SH Bair Side-b (without using DJ) with Marth, and allow you to be more creative with Side-b and other specials in your aerial game.

 

[Kadano]

alright............ I swear to god I just grabbed a fox during the bounce of missed tech...... Is this possible, or am I imagining it?

He most likely started a back-roll away from you. The first 11 frames of it are vulnerable and can be grabbed.

@ Kadano @ Bones0
Do you guys feel like Melee made the right decision to restrict Specials (and Air dodge) from being able to be used during IASA frames in the air like DJ and Aerials can? Not having this restriction would allow SH Bair WL and SH Bair Side-b (without using DJ) with Marth, and allow you to be more creative with Side-b in your aerial game.

I personally do not and would prefer air dodge and specials to be applicable for interruption as well.
If this is an(other) attempt of yours to get me involved with PM, I have to disappoint you, though—I have tried PM very often and there are too many things about this game I dislike, so I do not have any motivation to spend time with it, except for the occasional hour of freeplay with friends.

 

[SpiderMad]

Glad you feel the same way, I can at least rationalize Specials require expending your DJ given they're special. and with Air dodge probably because of it's extreme nature (and animation-wise) they didn't want to allow it to cancel out commitment on attacks and likewise to cancel Tumble and such either (which all got reverted in Brawl).

Well PM made/makes me aware of these different design choices/changes. I appreciate learning any of the stuff you feel is wrong with it, and I wouldn't want you to get "involved" more than just caring somewhat: given its constantly large changes each version throwing away a lot of people's work/findings.

 

[Bones0]

I'm not sure I like the idea of airdodging in IASA frames, but I don't see the harm in allowing specials. Then again, there might be some broken combos that become possible with specials during IASA. Idk which attacks are IASA heavy enough for that to matter (I just do everything by feel at this point), but just as an example, if Puff's fair had really good IASA, she might be able to light fair into rest or something ridiculous. Something like Sheik nair OoS -> needles could also be annoying. Falco wouldn't have much use for it, and all of Marth's moves act like normals anyway except counter, so it's not surprising something like fair -> B-move seems like a non-issue. I'd have to test some characters and see what options open up, but I don't think it'd make a huge difference either way. Airdodging would be the main problem I have with it. Truth be told, if I unknowingly played on a modded version of Melee with someone who was accustomed to interrupting with everything, I probably wouldn't even notice until they do an aerial -> airdodge.

 

[SpiderMad]

Well yeah that's how it is in PM (from Brawl?) and nobody notices or mentions it. But once you do recognize, you can totally feel it when switching between the games.

Like your Sheik's Nair example: you get to shave off 6 frames (unless you were to DJ first in Melee). I'm not sure if you guessed or not it having decent IASA.
http://smashboards.com/threads/sheik-hitboxes-and-frame-data.299984/

Jiggly's Bair would (/does in PM) shave off 8 frames to go into Pound/Rest (though one of her weaker DJs into Pound pry works fine)
http://smashboards.com/threads/jigglypuff-hitbox-and-frame-data-nsf56k.303202/#post-17168156

One thing is since everything now can work through IASA, there's even less need to have muscle memory for when the move's animation ACTUALLY ends. Good or bad idk; but yeah.

 

[Dart!]

Is the spacing for shino stall edgehog dash pivot grab far enough away where you can stand in position and not get faired from ledge? I want to use it this week but if I'm too close my opponent loves to fair me from edge then retreat back to ledge.[quote/] I never got an answer to this

 

[MookieRah]

Has anyone thought to do frame analysis on Marth's options against the spacies (or other high tiered characters) after you downthrow them so they are near the ledge? I'm thinking it's more than possible to create a flowchart that could cover most options quite optimally from reaction just from my own experiences with trying it out.

 

[dude it's raining]

Has anyone thought to do frame analysis on Marth's options against the spacies (or other high tiered characters) after you downthrow them so they are near the ledge? I'm thinking it's more than possible to create a flowchart that could cover most options quite optimally from reaction just from my own experiences with trying it out.

M2K has mentioned in his match analysis about frame perfect Dtilt allowing you to also cover tech roll in with a pivot grab.

 

[Bones0]

Good players just bounce onto the ledge from throws like that these days. I guess it might be helpful to know if you can hit them before they grab the ledge or not. I feel like it is only possible vs. really light characters like Puff. I never get dtilted going for bounces off of throws.

 

[1MachGO]

Yeah, the cancelling of hitstun might hinder Marth's edge tech chase potential. I could still see it being viable at certain positions/percents if it assures the opponent ends up offstage though.

 

[Vivec]

What's the window to hit fox and/or falco out of their firefox/firebird recoveries as Marth? What distance would this translate into (vertically/horizontally) before it is no longer possible to hit them without reading that they will use their up-b? Say Fox or Falco choose to angle their recovery to the ledge. How many frames does one have to react to the angles of the firebird/firefox? I'm specifically curious what the minimum distance (and the leniency of the window) needed to up-b for fox and falco in order for a player to move from the stage to the ledge quick enough to catch a firebird or firefox without getting burned?

 

[dude it's raining]

What is the best way to input a Utilt? I can do it if I hold slightly up before pressing A, but when I try to do it on reaction, I often end up jumping or Usmashing.

 

[MookieRah]

Good players just bounce onto the ledge from throws like that these days. I guess it might be helpful to know if you can hit them before they grab the ledge or not. I feel like it is only possible vs. really light characters like Puff. I never get dtilted going for bounces off of throws.

I'm actually not talking about a downthrow that puts your opponent all the way to the ledge. I'm talking about a downthrow that puts them relatively close to the ledge, giving them very little space to do much of anything, and limits what they can do.

Also, I knew about the dtilt to pivot regrab, it was something that I had in my Marth compilation guide I've been practicing that for a while now. I feel though, that it would be very possible to optimise this further than simply downthrow > dtilt > pivot regrab every time. For example if the post throw positioning is just right it seems it would be possible to land f-smashes against someone doesn't tech, techs in place, and tech rolls away fairly easily. Hell, if they are far enough away you can still catch them in a fsmash mid roll towards you, but that is hardly ideal/optimised. Also, tech roll to ledge would be a guaranteed tipper every single time (not saying people would be dumb enough to do that, but if they didn't know about this setup they might do it to "mix things up").

Basically, if a player could determine the subtle differences between all other tech situations and tech roll forward in time, you should be able to further optimise Marth's options further to include fsmash when it would be better than a d-tilt. I don't know if it's possible to do that though, but if it was it would be a good thing to do, because, at worst, you will be able to rack up lots of damage and put your opponent in a really bad off-stage position, and at best possibly close out a stock early with a tipper.

 

[Berble]

@ Kadano do you know the percents/di necessary for Marths throw into pivot fsmash combo on Puff? I saw PewPewU using dthrow pivot tipper on hbox at apex but im curious what the actual guaranteed setups are and if it works on other floaties.

 

[Berble]

Good players just bounce onto the ledge from throws like that these days. I guess it might be helpful to know if you can hit them before they grab the ledge or not. I feel like it is only possible vs. really light characters like Puff. I never get dtilted going for bounces off of throws.

does bouncing on to the ledge mean di towards the ledge and miss the tech so you edge cancel and grab the ledge?

 

[MookieRah]

does bouncing on to the ledge mean di towards the ledge and miss the tech so you edge cancel and grab the ledge?

It would be DI'ing towards the ledge, but I'm not sure that you'd get enough bounce off of not teching to actually cause you to ledge cancel or bounce so that you fall off the ledge unless you are pretty much right on top of it. Also, if you are familiar with the 20XX hackpack, comps will actually do this while still teching all the time, so I don't know how much the bounce has to do with it. Then again, the downthrow animation forces the opponent to bounce off the ground which causes the knockback, and could be referring to that.

In either case, it's mostly irrelevant to what I'm asking about because I'm not talking about performing downthrow's in which this is a possibility.

Also, in an attempt to find times in which PP used a dthrow to set up this kind of positioning I stumbled on this in the very first place I looked.
https://www.youtube.com/watch?v=Gv74JXJBFwk#t=32

Leffen attempts to bounce off the ledge for a grab as stated by those in this thread and still eats a dtilt. Perhaps Leffen could have fastfallen at some point, but if not then going for the ledge in this way is dangerous for Fox. Not to mention if this was the default DI then it could also be abused with potential uthrow followups more than likely.

 

[Binx]

@ Kadano do you know the percents/di necessary for Marths throw into pivot fsmash combo on Puff? I saw PewPewU using dthrow pivot tipper on hbox at apex but im curious what the actual guaranteed setups are and if it works on other floaties.

Its in the first post or at least a link to it.

 

[Bones0]

does bouncing on to the ledge mean di towards the ledge and miss the tech so you edge cancel and grab the ledge?

A bounce is simply your character's MT or TIP animation being cancelled because momentum carried you off of the initial surface you came into contact with.

It would be DI'ing towards the ledge, but I'm not sure that you'd get enough bounce off of not teching to actually cause you to ledge cancel or bounce so that you fall off the ledge unless you are pretty much right on top of it. Also, if you are familiar with the 20XX hackpack, comps will actually do this while still teching all the time, so I don't know how much the bounce has to do with it. Then again, the downthrow animation forces the opponent to bounce off the ground which causes the knockback, and could be referring to that.

In either case, it's mostly irrelevant to what I'm asking about because I'm not talking about performing downthrow's in which this is a possibility.

Also, in an attempt to find times in which PP used a dthrow to set up this kind of positioning I stumbled on this in the very first place I looked.
https://www.youtube.com/watch?v=Gv74JXJBFwk#t=32

Leffen attempts to bounce off the ledge for a grab as stated by those in this thread and still eats a dtilt. Perhaps Leffen could have fastfallen at some point, but if not then going for the ledge in this way is dangerous for Fox. Not to mention if this was the default DI then it could also be abused with potential uthrow followups more than likely.

Leffen's DI caused him to land very far from the ledge, and thus it took him longer than normal to grab the ledge. I'm quite certain he could have bounced closer to the ledge and FFed to grab it better than in the video. It's also worth noting that character weight will affect how easy it is to get the dtilt. Fox is going to be way more susceptible to getting dtilted on bounce attempts than Falcon or Samus would because you'll still be in the throw animation by the time those characters are FFing to the ledge.

Bouncing is based around your momentum and traction, which is why a Luigi at 100% doing a TIP by the ledge is going to have much more trouble grabbing the ledge than a Peach at 0%. The catch, of course, is that Peach will have a much harder time actually getting the bounce because she has less KB/momentum and less traction. She'll have to DI much more precisely to bounce out of her TIP/MT than Luigi would.

 

[EmuKiller]

I saw a gif of marth ledge hopping and maintaining invulnerability into the first 1-2 frames of his shield.

I'm guessing someone already asked about this, but Is there any info on the frame by frame inputs necessary to do this, and if it's stage dependent?

 

[Kadano]

Has anyone thought to do frame analysis on Marth's options against the spacies (or other high tiered characters) after you downthrow them so they are near the ledge? I'm thinking it's more than possible to create a flowchart that could cover most options quite optimally from reaction just from my own experiences with trying it out.

Things like this may seem easy to analyze, but there’s actually quite a lot to it in terms of variables you have to consider.

Variables:
x = distance from ledge (measured in mu (Magus units / Melee units), rounded to one decimal place)
p = percent (Fox’s damage before unstale throw)
y = TDI (trajectory directional influence) input (+1 yields the highest flying curve apex, −1 the lowest; not accurate, only estimated values)
t = amount of frames spent in Turn before starting the dtilt (not completely independent because its optimal value can mostly be told from x, p and y by evaluating the dtilt and move results below)
Results:
dtilt = which hitbubble of your dtilt connects (3 = sweetspot, 0 = blade, 1&2 = sourspot, no = no collision)
move = how far you have to move after the dtilt for the grab to connect

Thus, we have four mostly independent variables. Because the computer vision tools available to me here are limited to three dimensions (x, y and color), it is not possible for me to map out 5+ dimensions (four variables and at least one output) and display them here. (I’m not even considering the amount of time it would take to test every single combination.)
In other words, this micro-situation is too complex to do a complete frame analysis about, so I will eliminate p by setting it equal to 0 and only look at the “optimal” amount of t for a given DI so that I can create a table that is still both easy to understand intuitively and informative.

Right now, the table is valid for NTSC Fox. Falco is vastly different due to his stupidly long techroll, so don’t assume these values are true for him as well.

x|p|y|t|dtilt|move|comment
11|0|+1|-|-|-|closest distance from ledge where Fox will still land on the stage for +1 TDI
17.6|0|0|-|-|-|closest distance from ledge where Fox will still land on the stage for 0 TDI
16.5|0|−1|-|-|-|closest distance from ledge where Fox will still land on the stage for −1 TDI
24.6|0|0|-|-|-|closest distance from ledge where Fox will not slide off before dtilt hits for 0 TDI
35.2|0|−1|-|-|-|closest distance from ledge where Fox will not slide off before dtilt hits for −1 TDI

Now before I add more information to this table, we can already see that Marth needs to stand very far from the ledge to prevent Fox from doing missed tech to slide off, which evades both dtilt range and grab (obviously). It is, however, possible to catch it with a tipper forward smash, but only for certain distances.
Additionally, fsmash only connects when done immediately (assuming Fox DI’s perfectly and attempts to fast fall to the ledge), so considering it as well downgrades the strategy as a whole from a purely flowchart / reaction thing to a guessing thing.

When we only look at the flowchart part of it, it’s completely beaten by Fox’s missed tech → slide off → fast fall option. In other words, it only works on opponents who don’t know their options. So putting in more work into the table only makes sense when we include the instant fsmash option and accept that Marth needs to predict and cannot rely exclusively on reaction / flowchart execution.

I saw a gif of marth ledge hopping and maintaining invulnerability into the first 1-2 frames of his shield.

I'm guessing someone already asked about this, but Is there any info on the frame by frame inputs necessary to do this, and if it's stage dependent?

4 frames grounded actionable ledge intangibility is max for Marth, inputs are straightforward (drop without fastfall, jump as soon as possible and hold forward, airdodge diagonally as soon as you rise above stage level). No notable stage differences.

What you described is probably the no-impact land mentioned in the linked table. It’s worse than ledgedash concerning galint (1 frame galint is max), though. You need to fast fall for two frames before jumping for that.

 

[1MachGO]

Things like this may seem easy to analyze, but there’s actually quite a lot to it in terms of variables you have to consider.

Variables:
x = distance from ledge (measured in mu (Magus units / Melee units), rounded to one decimal place)
p = percent (Fox’s damage before unstale throw)
y = TDI (trajectory directional influence) input (+1 yields the highest flying curve apex, −1 the lowest; not accurate, only estimated values)
t = amount of frames spent in Turn before starting the dtilt (not completely independent because its optimal value can mostly be told from x, p and y by evaluating the dtilt and move results below)
Results:
dtilt = which hitbubble of your dtilt connects (3 = sweetspot, 0 = blade, 1&2 = sourspot, no = no collision)
move = how far you have to move after the dtilt for the grab to connect

Thus, we have four mostly independent variables. Because the computer vision tools available to me here are limited to three dimensions (x, y and color), it is not possible for me to map out 5+ dimensions (four variables and at least one output) and display them here. (I’m not even considering the amount of time it would take to test every single combination.)
In other words, this micro-situation is too complex to do a complete frame analysis about, so I will eliminate p by setting it equal to 0 and only look at the “optimal” amount of t for a given DI so that I can create a table that is still both easy to understand intuitively and informative.

Right now, the table is valid for NTSC Fox. Falco is vastly different due to his stupidly long techroll, so don’t assume these values are true for him as well.

x|p|y|t|dtilt|move|comment
11|0|+1|-|-|-|closest distance from ledge where Fox will still land on the stage for +1 TDI
17.6|0|0|-|-|-|closest distance from ledge where Fox will still land on the stage for 0 TDI
16.5|0|−1|-|-|-|closest distance from ledge where Fox will still land on the stage for −1 TDI
24.6|0|0|-|-|-|closest distance from ledge where Fox will not slide off before dtilt hits for 0 TDI
35.2|0|−1|-|-|-|closest distance from ledge where Fox will not slide off before dtilt hits for −1 TDI

Now before I add more information to this table, we can already see that Marth needs to stand very far from the ledge to prevent Fox from doing missed tech to slide off, which evades both dtilt range and grab (obviously). It is, however, possible to catch it with a tipper forward smash, but only for certain distances.
Additionally, fsmash only connects when done immediately (assuming Fox DI’s perfectly and attempts to fast fall to the ledge), so considering it as well downgrades the strategy as a whole from a purely flowchart / reaction thing to a guessing thing.

When we only look at the flowchart part of it, it’s completely beaten by Fox’s missed tech → slide off → fast fall option. In other words, it only works on opponents who don’t know their options. So putting in more work into the table only makes sense when we include the instant fsmash option and accept that Marth needs to predict and cannot rely exclusively on reaction / flowchart execution.

Welp, this option seems a lot worse than expected.

Is this mu you are speaking of the same metric as scmooblidon's mm? If that's true then you have to be FAR from the edges for this to even work at 0%.

I guess platform combos will remain the standard.

 

[MookieRah]

Thanks for all the info, now I know that I shouldn't focus on this scenario, at least not like I have been. I am thinking now that it's better to downthrow in order to simply put Fox/Falco in a really bad position and then react off of what they do afterwards, as opposed to trying to find a way to cover all their options near-instantly.

 

[Kadano]

Is this mu you are speaking of the same metric as scmooblidon's mm? If that's true then you have to be FAR from the edges for this to even work at 0%.

Yeah, it’s the same. Like I wrote before, when viewed conventionally, there is no spacing where it works at all if Fox uses his DI down and away → slide off to fastfall ledgegrab option. If you are close enough to the ledge for him to be in dtilt range, he will slide off safely. If you are far enough away from the ledge for sliding off to be impossible, he will be out of dtilt range.

Thanks for all the info, now I know that I shouldn't focus on this scenario, at least not like I have been. I am thinking now that it's better to downthrow in order to simply put Fox/Falco in a really bad position and then react off of what they do afterwards, as opposed to trying to find a way to cover all their options near-instantly.

Well, if your opponent can ledgedash consistently, they end up in an advantageous position whenever they are close enough to the ledge for the slide, I’d say. So when trying to do it like that, you should be aware of the range you have to be away from the ledge for this option to not exist.

 

[Signia]

@ Kadano

Sheik
Grab options against Sheik
Upthrow takes 40 frames against her.
0-8% – 2-3 frame window for Sheik’s jump escape timing (cannot be buffered)
12% – 2 frame window
13% – 1 frame window
17% – 2 frame window
21% – 1 frame window
22-30% – utilt is guaranteed, you need to turn if she DIs behind you
35-41% – utilt is guaranteed on neutral DI only, otherwise it will whiff.
32-45% – If she DIs behind you, tipper fsmash will miss just barely (she needs to hit a one frame window for the jump to avoid it). If she DIs behind you or in front of you, short hop fair will tipper her if done frame perfect (dash 1 frame, jump for one frame, then input the fair).
29-57% – On neutral DI, short hop uair will always hit her.
50-70% – On neutral DI, full jump uair will always connect, you need to delay the uair by one frame every 6% or so.
70-86% – On neutral DI, full jump → second jump uair will always connect.

From 90% onwards, it seems that there are no guaranteed followups. All percentages are given as before hit, although that’s not like Melee calculates it.

This has been very helpful. I have few questions, though:

Why is there a gap from 31-34%? is there really no followup? How about 9-11%? 14-16%? 18-20%?

Also, in places where utilt is guaranteed, is it a one-frame timing (for Marth)?[/quote]

 

[Kadano]

@ Kadano

This has been very helpful. I have few questions, though:

Why is there a gap from 31-34%? is there really no followup? How about 9-11%? 14-16%? 18-20%?

Also, in places where utilt is guaranteed, is it a one-frame timing (for Marth)?

[/quote]
The gaps are just ranges that I did not test specifically to be able to give reliable information. (Doing these things is rather time-consuming, so I settled for approximate ranges and not for precision down to one percent. In-game, staling matters as well, and I don’t think anyone memorizes his staling queue, so 5 percent ranges are sufficiently precise, I think.

Yeah, it’s a one frame link usually. Probably two frames for a few percent ranges where hitstun increases, but knockback not so much that she’d escape utilt range within that one frame of leniency.

 

[BTmoney]

For uthrow on spacies (specifically falco) I can never seem to get the CG right in the 20XX pack.
If I u-tilt at about 32% and regrab and uthrow again and the CPU hard DI's away all I can seem to get it f tilt, dash fair, and non tipper f smash.

I think you're supposed to take 2 steps and u tilt around 50ish maybe a pinch higher but I can never get that to connect. Are my percents just wrong? I was pretty confident about doing your first u-tilt at 32%

 

[tonic]

just regrab if they DI away <60ish percent.

 

[1MachGO]

For uthrow on spacies (specifically falco) I can never seem to get the CG right in the 20XX pack.
If I u-tilt at about 32% and regrab and uthrow again and the CPU hard DI's away all I can seem to get it f tilt, dash fair, and non tipper f smash.

I think you're supposed to take 2 steps and u tilt around 50ish maybe a pinch higher but I can never get that to connect. Are my percents just wrong? I was pretty confident about doing your first u-tilt at 32%

Whats wrong with uair? Regrab should also work if they hard DI away.

However, if you really want to utilt, it might be possible out of a WD. Uthrow at 50% gives Marth 21 frames to follow up on Fox and 20 on Falco. This means WD utilt BARELY won't combo (14 frames for WD, 8+/- for utilt). I imagine the height is low enough that Fox/Falco would still get hit by utilt even if they jumped. This means they are forced to take the hit unless they want to be juggled into a tipper without a DJ. @ Kadano are there any implications for this?

 

[Dart!]

Does port priority affect any of the chaingrabs?

 

[Kadano]

Does port priority affect any of the chaingrabs?

Yes.

 

[1MachGO]

Yes.

How so?

 

[AprilShaw]

Alright I have some questions concerning Luigi...

I heard that it's possible to always dtilt/fsmash Luigi's up B, as in he can't sweetspot around it. Is this true, and if so, is it possible to react to the sound of the up B and still hit him? Would dsmash work as well?

EDIT: Found a way to test this on my own, found a spacing where you can't hit him with anything from on stage that isn't an aerial. *sadface*

EDIT 2: At that spacing it is possible to grab the edge before him by wavedashing, so I suppose you could react to him going super deep and just grab the ledge... really tight window though

The other 2 aren't nearly as specific but first, what throws do people like using at very low (<10) and very high (>100) percents? Uthrow seems to be the obvious choice but at low percents I can't get a follow up consistently (could easily be me sucking though) and at high percents he can use platforms to make it a lot more difficult. And finally, does anyone know of reliable kill setups against him besides getting him to side B on a platform and fsmashing it? I'm debating switching to Sheik completely for the matchup just cause killing off of throws is sooo nice, lol

Thanks for the help

 

[MookieRah]

Against floaty characters in which throws do not grant any guaranteed follow up, I personally find it best to throw them towards the ledge. My entire strategy against them (and pretty much every other character) is to pin them to the stage so they have as few options as possible.

Also, Marth wrecks Luigi's face so there is absolutely no need to switch characters. Luigi is entirely dependent upon the ground and platforms to have any sort of horizontal movement. Dtilt wrecks his face, and as long as you are careful about his various vectors of attacks from platforms, you should be alright, as platforms are a bad place to be against Marth anyways.

If you just play a clean, solid game in which you mind your spacing you shouldn't have any trouble, even if he manages to live till very high percents. Just don't get flustered.

 

[Bones0]

Alright I have some questions concerning Luigi...

I heard that it's possible to always dtilt/fsmash Luigi's up B, as in he can't sweetspot around it. Is this true, and if so, is it possible to react to the sound of the up B and still hit him? Would dsmash work as well?

EDIT: Found a way to test this on my own, found a spacing where you can't hit him with anything from on stage that isn't an aerial. *sadface*

EDIT 2: At that spacing it is possible to grab the edge before him by wavedashing, so I suppose you could react to him going super deep and just grab the ledge... really tight window though

The other 2 aren't nearly as specific but first, what throws do people like using at very low (<10) and very high (>100) percents? Uthrow seems to be the obvious choice but at low percents I can't get a follow up consistently (could easily be me sucking though) and at high percents he can use platforms to make it a lot more difficult. And finally, does anyone know of reliable kill setups against him besides getting him to side B on a platform and fsmashing it? I'm debating switching to Sheik completely for the matchup just cause killing off of throws is sooo nice, lol

Thanks for the help

If you want to kill off of throws, you're playing the wrong character.

 

[BTmoney]

Anybody look at @gravy's quarter circle empty pivots? They are pretty amazing (i.e. this means you can d-tilt out of a dash dance or at least a dash back and it's not impossibly hard). I could get a lot after only a few minutes of practice, you really just have to get the motion down. It's a little awkward. And gravy also said he's ~85-90% consistent on pivoting on demand so this seems very possible to master for at least dash back->pivot d-tilt.

I do not think this method will work for an advancing pivot d-tilt (because it turns you against the direction of your dash) but it should work for implementing your pivot F smashes out of throws since you can f smash in either direction without need to turn around. I find getting those advancing pivots to be very challenging to land, especially to the left.

http://smashboards.com/threads/this-is-how-to-move-good-and-do-other-things-good-too.389487/
ctrl + f "quarter"

He calls it a quarter circle but I think of it as a triangular motion (imagine the bottom left or right 1/4th of your analog stick's range of motion, make a right triangle inside that arc. When I do this I don't really rotate, I go from my dash to -> straight to the opposite side -> straight down very quickly and firmly. It's kind of like a moonwalking motion but on that 1 frame you are standing up and down you start inputing a crouch to make you stand in place.

The trick is to move from horizontal to straight down vertically in 1 frame. I find when I try to "rotate" to make a 1/4th circle I do it pretty slow but when I think of a jagged triangular motion I get it much more often.

To get the motion down he recommends trying to do a pivot down smash with just the analog stick and A. It's the same motion and it makes a lot more sense once you do it once or twice. I find this to be a little harder than double shining (grounded) and notably harder shield dropping. At least right now, but if you looked at the frame data for shield dropping it looked like it'd be impossible to ever master but now it's almost mindless for me.

 

[AirFair]

Yo @ Kadano can I ask if certain combos are guaranteed here?