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

[hectohertz]

Is there a way to allow moves to stale in develop / debug mode?

P.S. still looking for answers to this:

its a % threshold. i don't remember what though. i think there's a magus post explaining it somewhere

 

[Meru]

So how do JC's work frame-wise? For marth to do a JC grab would it add 4 frames of jumpsquat to the startup making its active frames effective on the 11th frame instead of the 7th? Would this not make dash grab just as/more effective (because of length) in a chaingrab situation where you have to chase down somebody who has done full DI?

 

[Bones0]

So how do JC's work frame-wise? For marth to do a JC grab would it add 4 frames of jumpsquat to the startup making its active frames effective on the 11th frame instead of the 7th? Would this not make dash grab just as/more effective (because of length) in a chaingrab situation where you have to chase down somebody who has done full DI?

You only need to go through 1 frame of jumpsquat in order to get a standing grab. Pressing Z to grab on the last frame of jumpsquat would actually result in an aerial because you are airborne frame 5.

 

[Jim Morrison]

Yo boys, how does light shielding differ from normal shielding, especially frame advantage-wise? Do you actually get longer shield-stun from it or does it only feel like it?

 

[tm]

Yo boys, how does light shielding differ from normal shielding, especially frame advantage-wise? Do you actually get longer shield-stun from it or does it only feel like it?

Yes you do. 1.5x if it's fully lightshielded, and I believe it scales linearly.

 

[Awstintacious]

You are a god @ Kadano , if only 1/5 as many of us gave 1/3 of a ****, we'd have an awesome time.

 

[Endeby]

It has been known for a few years that PAL Marth has an maximum aerial velocity of 0.85 as opposed to 0.90 in NTSC. Has the practical implications of this ever been discussed? I honestly only see people chat about the Dair change, so I'm starting to wonder if this isn't that much of a big deal.

I figure it most likely make aerial follow-ups slightly harder in PAL compared to NTSC and change other aspects of his aerial game as well. Like when he's trying to recover from the upper corner back to the stage? I never really see people mention it.

Maybe it would be better of me to make a new thread for this, but someone like @ Kadano , who know both versions of Marth, might have something on this?

 

[hectohertz]

It has been known for a few years that PAL Marth has an maximum aerial velocity of 0.85 as opposed to 0.90 in NTSC. Has the practical implications of this ever been discussed? I honestly only see people chat about the Dair change, so I'm starting to wonder if this isn't that much of a big deal.

I figure it most likely make aerial follow-ups slightly harder in PAL compared to NTSC and change other aspects of his aerial game as well. Like when he's trying to recover from the upper corner back to the stage? I never really see people mention it.

Maybe it would be better of me to make a new thread for this, but someone like @ Kadano , who know both versions of Marth, might have something on this?

thats crazy, literally a minute ago i was reading the changes between PAL and NTSC and noticed this too. was coming here to post about it.


weirddd

 

[Kadano]

@ Endeby @ hectohertz

It seems to be very minor.

 

[AprilShaw]

Never knew there was a difference in the stock icons between NTSC/PAL o.o

 

[Endeby]

Amazing! i guess I can practice my Marth in both PAL and NTSC .

 

[Purpletuce]

Kadano, is there any chance you can find some numbers for me? I'm curious about the relative sizes of the stages (ledge to ledge width), the platform heights, and blastzone distances(distance a character has to travel from center stage to die off of the top or sides).

I'd be interested in making a graphical representation of the various stage properties, and don't know where else to get the numbers to start with.

 

[AprilShaw]

Might be worth asking Achilles/Internetexplorer for that one, too. You can see/change the platform heights on BF in the 20XX pack

 

[Kadano]

Kadano, is there any chance you can find some numbers for me? I'm curious about the relative sizes of the stages (ledge to ledge width), the platform heights, and blastzone distances(distance a character has to travel from center stage to die off of the top or sides).

I'd be interested in making a graphical representation of the various stage properties, and don't know where else to get the numbers to start with.

See post #1 → Hitboxes[…].xlsx. One of its sheet is called “stages” and has all that data. I think it was Strong Bad who collected it, but I don’t remember where I found it exactly.

 

[Purpletuce]

Amazing. I didn't think to check the hitboxes spreadsheet. Thanks to you for forwarding me that info, and thanks to Strong Bad for generating it.

Edit: Data is missing for PS, any chance you know those numbers/they are somewhere else?

 

[DJ _ICE]

https://www.youtube.com/watch?v=Alh64TmBz6U#t=71 Is rage quitting this fast really possible? The pause screen doesn't even show up

 

[Tee ay eye]

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

kadano pls

 

[woodsta]

I'm gonna guess that the slanted underside of battlefield is forcing fox downwards a little bit at that angle and his rear grab box overlaps during that short time before he slides down too far to grab it

 

[Kadano]

https://www.youtube.com/watch?v=Alh64TmBz6U#t=71 Is rage quitting this fast really possible? The pause screen doesn't even show up

I don’t know how I could find it out. I was not able to do it anywhere near as fast with Dolphin’s TAS inputs and frame advance, but I feel it was somehow Dolphin’s fault. It’s not possible to test this using Melee’s develop frame advance, and I really don’t know any other ways.

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

kadano pls

That happens extremely rarely. What happens has to be the same as in Magus420’s video (which is exactly what @ woodsta wrote), but for Fox to have the correct spacing is next to impossible. I tested the angles for 4 different positions, and for none of them any angle worked. I’m pretty sure it requires too precise positioning and inputs to ever be useful on purpose.

 

[Bones0]

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

kadano pls

Yeah, it's probably much like Falcon's in that on that stage it's possible to get pushed down from the nearby ceiling surface while also having only the one ledge grab box connect with the ledge area, which meets the conditions needed to grab an edge and so you do.

If it's pretty much exactly like Falcon's you need to hit where the wall and ceiling join just right (you can see this part of the stage in that video) so that you interact with both surfaces on the same frame: getting pushed down by hitting the ceiling for the downward movement requirement, and pushed outward by the wall enough so that only the front ledge grab box is making contact with the ledge (if both the front and rear grab boxes make contact the game prevents you from grabbing it). Being too far into the stage so that they both connect is how you get "Battlefielded".

And yeah, I tried to replicate it manually for like an hour and couldn't get it.

 

[schmooblidon]

And yeah, I tried to replicate it manually for like an hour and couldn't get it.

lolol same here. dis tech mang

 

[Kadano]

I tested the angles for 4 different positions, and for none of them any angle worked. I’m pretty sure it requires too precise positioning and inputs to ever be useful on purpose.

And yeah, I tried to replicate it manually for like an hour and couldn't get it.

lolol same here. dis tech mang

Only @Magus420 or @ajp_anton could do this, I think. Magus would work his magic to reverse-engineer the Melee code and calculate the position and angle Fox needs to recover at, and ajp_anton would write scripts that make Dolphin test every one of them until one of them works.

Sadly, they both seem to not do this kind of stuff any longer.

 

[Bones0]

I think CPUs always go at a 45, or at least they do in every example of the instant grabbing I've seen. If there is an angle that makes it more likely to occur, maybe it'd actually be possible. I think doing the one that goes straight up is actually easier. I remember I was able to get it ~10% of the time or something. The problem is obviously that you just die the other 90%.

 

[SpiderMad]

I don’t know how I could find it out. I was not able to do it anywhere near as fast with Dolphin’s TAS inputs and frame advance, but I feel it was somehow Dolphin’s fault. It’s not possible to test this using Melee’s develop frame advance, and I really don’t know any other ways.


That happens extremely rarely. What happens has to be the same as in Magus420’s video (which is exactly what @ woodsta wrote), but for Fox to have the correct spacing is next to impossible. I tested the angles for 4 different positions, and for none of them any angle worked. I’m pretty sure it requires too precise positioning and inputs to ever be useful on purpose.

I heard you had to press Pause twice or something to get the quick one. Or wait, it might be hold Pause and release it or something like that.

I love how there's always one way to recover from the impossible that we can't even replicate, like there's magic in the game waiting for when someone needs to make the clutch come-back in tournament

@ Bones0 Is there any downside for Full pressing a shield (essentially triggering a Powershield) versus full shielding with the analog (besides not wanting the powershield for some reason?). I was thinking it's possible that you're inputting tech and could screw up your tech window if you somehow got hit before the shield came out; and that being the possible reason they didn't just assign Powershield to full shield (as well as giving you the option to full shield without powershielding).

 

[He-Man1]

https://www.youtube.com/watch?v=Alh64TmBz6U#t=71 Is rage quitting this fast really possible? The pause screen doesn't even show up

I don’t know how I could find it out. I was not able to do it anywhere near as fast with Dolphin’s TAS inputs and frame advance, but I feel it was somehow Dolphin’s fault. It’s not possible to test this using Melee’s develop frame advance, and I really don’t know any other ways.

The reason for this is actually that the capture dropped frames while they were recording lol. It's pretty common in HMW's videos and seems to happen more frequently when the commentators are yelling really loudly.

 

[Tee ay eye]

kadano, could you please explain a few things about DI to me?

1. how does shy guy DI work? i don't really understand it in terms of game mechanics like SDI/ASDI and hitboxes and whatnot
2. how exactly does.... slide DI (can't think of a better term) work? like, whenever you DI in a way that allows you to ledgecancel off an edge into no lag. is it just a matter of getting good DI on the hit/throw so the momentum from *that* makes you slide off? or is it more about the inputs you're doing when your character is actually interacting with the edge you want to cancel off of?

 

[tauKhan]

2. how exactly does.... slide DI (can't think of a better term) work? like, whenever you DI in a way that allows you to ledgecancel off an edge into no lag. is it just a matter of getting good DI on the hit/throw so the momentum from *that* makes you slide off?

Pretty much just that. There's really no mystery to it: Colliding into ground doesn't stop all of your horizontal momentum, and if you fall of edge, hitstun will cancel from the ground -> air transition. So if you want to slide off the edge to cancel stun, your goals are to 1: collide with the ground 2: Have enough horizontal momentum to carry you off an edge.

Sometimes the goals are kind of contradictory: If you are sent upwards at relatively low speed, you need to di to side to get any horizontal momentum at all. But to collide to ground you would need to asdi down. The best possible di in this case would be DSDI, toward edge with controlstick and down ASDI with c-stick. I do this often if I'm about to get hit on platform, it works great against marth uairs for example.

 

[Kadano]

1. how does shy guy DI work? i don't really understand it in terms of game mechanics like SDI/ASDI and hitboxes and whatnot

When a character is thrown, a hitbubble at his chest is active for a certain amount of time, which seems to depend on launch speed. For uthrows, it lasts more or less until the apex of your flying curve.
So if you are thrown into a shy guy, your hitbubble connects with their hurtbubble. Just like when you hit them with an attack—except that you are in hitstun, ie a Damage* animation. Whenever you are in any of these and you undergo hitlag, you can apply SDI, ASDI and TDI (this is what usually happens every time you get hit, which is why the game checks for that). For every character’s throw, this hitbubble does 3 damage. 3 damage causes 4 frames of hitlag. When you hit another shy guy, you go into hitlag again and are still in hitstun, so you can alter your trajectory with TDI again.

2. how exactly does.... slide DI (can't think of a better term) work? like, whenever you DI in a way that allows you to ledgecancel off an edge into no lag. is it just a matter of getting good DI on the hit/throw so the momentum from *that* makes you slide off? or is it more about the inputs you're doing when your character is actually interacting with the edge you want to cancel off of?

Adding to what tauKhan wrote, I created a video that explains all of that visually: https://www.youtube.com/watch?v=ZkdPVUlrSOo&t=6m30s

 

[tonic]

Is marth's optimal ledgedash setup the same on every stage? Kadano mentioned that the earliest frame marth could airdodge downwards was frame 13, but does that actually yield the most invincibility? I think marth mains need to take as much advantage as we can seeing as even our perfect ledgedash grants such short intangibility.

 

[hectohertz]

Is marth's optimal ledgedash setup the same on every stage? Kadano mentioned that the earliest frame marth could airdodge downwards was frame 13, but does that actually yield the most invincibility? I think marth mains need to take as much advantage as we can seeing as even our perfect ledgedash grants such short intangibility.

kadano: please correct me if wrong

but, since marth has no ecb manipulation, and optimal ledgedash should be drop, jump on first frame, then airdodge on the first frame your ecb is above the stage, airdodging on the earliest frame should *always* yield the most invincibility

now whether its frame 13 for all stages, that's an intersting question

 

[Kadano]

@ T tonic @ hectohertz it’s frame 13 for all legal stages—and, I suspect, all other stages as well, but I only tested the legal ones.

 

[Bones0]

@ T tonic @ hectohertz it’s frame 13 for all legal stages—and, I suspect, all other stages as well, but I only tested the legal ones.

Does KJ64's lack of a wall affect anything? I'd guess that it wouldn't.

 

[Meru]

Is it theoretically possible to Amsah tech any attack? If not, and there's some sort of cutoff angle for being able to amsah tech, what is it?
Also, what's the equation for finding the amount of hitlag experienced by an attacker landing a hit on shield?
Finally, what frames can shiek grab the ledge after her upb's "poof"? I was wondering how to time the neutral getup from ledge to combat shiek's recovery. I've been punishing it pretty sub-optimally for some time now.

 

[Bones0]

Is it theoretically possible to Amsah tech any attack? If not, and there's some sort of cutoff angle for being able to amsah tech, what is it?
Also, what's the equation for finding the amount of hitlag experienced by an attacker landing a hit on shield?
Finally, what frames can shiek grab the ledge after her upb's "poof"? I was wondering how to time the neutral getup from ledge to combat shiek's recovery. I've been punishing it pretty sub-optimally for some time now.

First two questions: https://www.youtube.com/user/meleemechanics/videos

 

[1MachGO]

Hey kadano, at kill %, should marth sdi up against fox's uair if he misses uthrow DI or is sdi'ing behind fox usually the best choice?

 

[dude it's raining]

Hey Kadano, can you tell me about JC grabs? Like sometimes I slide more or less than I think, and I keep feeling like I should be able to control and predict it, but I can't. I imagine it has something to do with initial dash acceleration or something?

Also, thought I should let you know, I'm making a general guide partially inspired by yours: http://smashboards.com/guides/down-the-rabbit-hole.108/

 

[1MachGO]

@ dude it's raining

Marth's dash momentum will transfer into jump squat. This means that the more frames spent in jump squat (delaying your grab) the further you will slide.

 

[dude it's raining]

@ dude it's raining

Marth's dash momentum will transfer into jump squat. This means that the more frames spent in jump squat (delaying your grab) the further you will slide.

Thanks so much!

 

[hectohertz]

@ dude it's raining

Marth's dash momentum will transfer into jump squat. This means that the more frames spent in jump squat (delaying your grab) the further you will slide.

is this true for all characters?

i remember for Ganons CGs you needed to do perfect JC grabs, although perhaps that meant the latest frame in jumpsquat, not the earliest as i had assumed

 

[1MachGO]

is this true for all characters?

i remember for Ganons CGs you needed to do perfect JC grabs, although perhaps that meant the latest frame in jumpsquat, not the earliest as i had assumed

Not sure. Magus has written up a lot of Ganon stuff so he might have that kind of detail specified.

I imagine that it ultimately comes down to the dash/run speed and traction of a character. Marth has pretty high run speed and low traction so time spent in jump squat can create some meaningful distance. Ganon is on the lower end in both regards so its hard to say how much of an impact there is.