Doc Advanced Tech Frame Data Megathread

[Fortune]

Thanks for this thread.

 

[MudkipUniverse]

I don't see why everyone hates fortune for what happened at "Smash @ Xanadu. It was the TO's fault for not making that pause an exception.

 

[Fortune]

He actually DID make the pause an exception. SG did not SD due to pausing. People were just angry that I tried to do it. I had a feeling the stream would take it too seriously, which is the main reason I ended up asking for the SD. I did not expect it to become a week-long event, though.

 

[1MachGO]

I got a couple questions for ya @ schmooblidon

I was looking at the doc hitboxes and frame data thread and was wondering if you could do a comparison in regards to how far certain hitboxes go below the ledge.

Specifically:
Does angled down ftilt go lower than dtilt?
How low does jab 3 go? The hitbox gets interpolated a lot and looks like it goes pretty darn low; I'm wondering if Mario's jab string has some application in regards to edge guarding Fox/Falco side-bs.

Lastly, could you do a comparison between Doc's run grab and stand grab ranges? The difference appears pretty significant.

Here is what Kadano did for Marth if you want to know what I mean by comparison: http://img836.imageshack.us/img836/8624/83m.png

 

[schmooblidon]

Unfortunately my motherboard died a few days ago, and I currently don't have the money for replacements. So I can't make any gif or pics at the moment. This was actually one of things I have planned to do, tonnes of other stuff too.

I can answer your questions though.
In order of lowest point to highest point:
Cape
Jab 3
Ftilt downward (only last hitbox frame) = Dtilt

The problem with jab 3 is that the low point of the hitbox is so close to doc, you have to stand right at the edge for it to be used as an edgeguard, still a valid option.

Dash>JC Grab is about 2/3 the range of running grab.

 

[Maple/DBK]

Damn, time for me to hit the lab with Mario! That Haxdash looks sick as hell to perform, especially for hand warmer sessions.

 

[Maple/DBK]

But like, can anyone tell me how to do the Haxdash? it looks really difficult to pull off.

 

[schmooblidon]

But like, can anyone tell me how to do the Haxdash? it looks really difficult to pull off.

Well, there are tonnes of slightly different setups that work. The problem is that they need a very tiny angle on the control stick making them pretty much impossible. You are welcome to try though, the general gist of it is:

Perfect ledgedrop
DJ towards stage
Airdodge diagonally downward into the stage and sliding off
Fastfall

For the longest time I thought using the stick diagonal notch made it impossible to haxdash with the most optimal setup giving 1 tangible frame. But I actually found yesterday that it is possible on yoshis due to the ledge's snapbox being higher then the actual ledge (because of the slant), albeit with a tiny margin for error.

Yoshis human-like haxdash
0 - CliffWait 0
1 - Drop
2 - Fastfall
8 - DoubleJump towards stage
9 - Hold towards stage
16 - Earliest time to stop holding towards stage*
25 - Airdodge (diagonal notch)
28 - Fastfall
30 - Grab ledge - Cliffcatch 1
37 - CliffWait 0

*becomes later the more you hold the airdodge angle before airdodging

This is like just barely do-able, and unless you are a robot, then something just for show. Holding towards the stage kinda sucks and makes it harder, the fastfall will only be performed automatically if you input the airdodge angle on the frame you airdodge, any earlier and you need to input another down movement.

I'll try to break down a human-like input:
(on the right ledge)
1 -
4 -
7 -
8 -
19 -
20 -
25 -

29 -
38 - etc

 

[Maple/DBK]

Well, there are tonnes of slightly different setups that work. The problem is that they need a very tiny angle on the control stick making them pretty much impossible. You are welcome to try though, the general gist of it is:

Perfect ledgedrop
DJ towards stage
Airdodge diagonally downward into the stage and sliding off
Fastfall

For the longest time I thought using the stick notch made it impossible to haxdash with the most optimal setup giving 1 intangible frame. But I actually found yesterday that it is possible on yoshis due to the ledge's snapbox being higher then the actualy ledge (because of the slant), albeit with a tiny margin for error.

Yoshis human-like haxdash
0 - CliffWait 0
1 - Drop
2 - Fastfall
8 - DoubleJump towards stage
9 - Hold towards stage
16 - Earliest time to stop holding towards stage*
25 - Airdodge (diagonal notch)
28 - Fastfall
30 - Grab ledge - Cliffcatch 1
37 - CliffWait 0

*becomes later the more you hold the airdodge angle before airdodging

This is like just barely do-able, and unless you are a robot, then something just for show. Holding towards the stage kinda sucks and makes it harder, the fastfall will only be performed automatically if you input the airdodge angle on the frame you airdodge, any earlier and you need to input another down movement.

I'll try to break down a human-like input:
(on the right ledge)
1 -
4 -
7 -
8 -
19 -
20 -
25 -

29 -
38 - etc

Damn, that really does sound difficult to pull off. Would it be that useful compared to like Hax$ when he mained Captain Falcon or it would just be useful for showing off/hand warmers?

 

[schmooblidon]

It would be great if you could actually pull it off repeatedly. But for it to be useful it must be perfect, which is like 6 frame perfect inputs per second. Falcon and many others have much much easier and more lenient haxdashes.

 

[Maple/DBK]

It would be great if you could actually pull it off repeatedly. But for it to be useful it must be perfect, which is like 6 frame perfect inputs per second. Falcon and many others have much much easier and more lenient haxdashes.

I'll try that, thanks for the info! great thread, by the way. The AIA tech seems to be quite useful in competitve play.

 

[#HBC | Mac]

you can pill on the 10th frame airborne and still waveland (only on the very last frame, no fast fall)

full jump lasts 52 frames and a pill lasts 43 frames

so does that mean:
if you pill on the first frame airborne and don't fast fall, then you have 9 (or 10) frames to do an aerial? so that should mean that you can do a fullhop pill -> aerial for any of doc's aerials that have an active hitbox out within 9 frames. so doc can do FH pill -> nair / bair / uair (but not dair and def not fair)

if you do a FH pill and double jump on the first actionable frame and pill again. How many frames do you have after the second pill to do an aerial?

edit: i wish i knew how to make hitbox gifs.. it would be really cool to see how much ground the cape hitbox covers when doing: full dash sh reverse cape

 

[#HBC | Mac]

how much does the airborne cape delay you in the air?

so a SH lasts 32 frames, how many frames are you airborne when you do a SH -> cape? Does it depend on your upwards momentum or something?

 

[schmooblidon]

full jump lasts 52 frames and a pill lasts 43 frames

so does that mean:
if you pill on the first frame airborne and don't fast fall, then you have 9 (or 10) frames to do an aerial? so that should mean that you can do a fullhop pill -> aerial for any of doc's aerials that have an active hitbox out within 9 frames. so doc can do FH pill -> nair / bair / uair (but not dair and def not fair)

if you do a FH pill and double jump on the first actionable frame and pill again. How many frames do you have after the second pill to do an aerial?

edit: i wish i knew how to make hitbox gifs.. it would be really cool to see how much ground the cape hitbox covers when doing: full dash sh reverse cape

There is something I didn't mention in that section, I don't think I actually knew about it when I wrote it honestly. Whenever you perform a move on the first frame airborne, you lose height and the total airtime is smaller. This is true for all characters, shorthop loses 1 frame, fullhop loses 2 frames.

So if you pill on the first frame without fastfall, you will be airborne for 50 frames. Then you also have to consider the aerial's ecb pattern, upair has big vertical movements, nair raises you up a tiny bit instantly, bair raises you a fair bit on the hitbox frames. Here is a quick table to show the windows, frame numbers are airborne frames.

Aerial | Starting frame | Hitbox frames | Total airborne frames
Upair | 44 | 2 | 48
| 45 | 1 | 48
| 46 | 1 | 49
| 47 | 0 | 49
Nair | 44 | 5 | 50
| 45 | 4 | 50
| 46 | 3 | 50
| 47 | 2 | 50
| 48 | 1 | 50
| 49 | 0 | 50
Bair | 44 | 3 | 51
| 45 | 2 | 51
| 46 | 1 | 51
| 47 | 0 | 50

I can tell you that if you were to pill on the second frame airborne, each aerial has a 1 frame greater window.
upair4, nair6, bair4

Case 1: FH pill first frame > DJ pill > no aerial, starting from the first DJ frame, you have 55 airborne frames.
Case 2: FH pill second frame > DJ pill > no aerial, you have 57 airborne frames.
So with a bit of maths you can guess that for case 1, upair has a 7 frame window, nair - 9, bair - 7. For case 2, upair - 9, nair - 11, bair - 9. Dair also comes into the equation and I can tell you that case 1, it has a 2 frame window, and case 2, a 4 frame window.

Ok so this is gunna be kinda hard to explain, but I'll give it my best. When you cape, the first 10 frames do nothing, and you will travel through the air like normal. On frame 11, your previous air velocity, acceleration, whatever don't matter anymore, you are now in a cape jump. It has it's own velocity and you will peak at frame 20 always. It's like a little doublejump.

The problem with working out how much it delays you in the air, is that it depends when you start it. The shorthop's velocity is much greater then the capejump's, so canceling it early will increase the total airtime far less then if you caped later. Shorthop peaks at frame 16, so if you cape on airborne frame 6, so that you cancel the shorthop at it's peak, you can reach the highest point possible. But this is not the maximum potential airtime. The higher you reach the more time gravity has to acceleration you downward. If you cape during the downfall, you will still have to travel both the shorthop height and the capejump height, but be able to have a lower average downward velocity.

I'll still throw some numbers at you though, as it follows a vague pattern.

Cape on airborne frame | Total air time
1 | 38
2 | 40
3 | 41
4 | 42
5 | 43
6 | 44
etc
16 | 52 (max)
22 | 52 (last frame to be able to capejump)

You can use that last result to roughly determine what a capejump from the ground would look like (20 airborne frames from frame 11).

 

[Kadano]

@ schmooblidon you might want to use table code for things like this, as smashboards doesn’t support tabulator breaks or multiple spaces. It’s not really accessible / properly documented, so I felt like telling you. (If you think this just clutters up the thread, just tell me; I can simply delete this post with my new mod powers.)

It’s used like this:

[table=head]Aerial|Starting frame|Hitbox frames|Total airborne frames
Upair|44|2|48
|45|1|48
|46|1|49
|47|0|49[/table]

produces:

Aerial|Starting frame|Hitbox frames|Total airborne frames
Upair|44|2|48
|45|1|48
|46|1|49
|47|0|49

 

[schmooblidon]

Oh man I thought it was way more complicated then that, cheers dude

 

[#HBC | Mac]

wow @ schmooblidon you spoil us

 

[#HBC | Mac]

good to know that how fast you do your cape when SHing will alter how long you stay in the air. I'm not sure what in-game applications this has though

If i'm understanding all this correctly, if you cape asap in a SH you probably won't be able to fit in any aerials after the cape, but you'd still be able to waveland, which might be useful for being speedy when gaining distance (vs falco laser or something) whereas if you're caping and want to follow up with some aerial, you want to start the cape a lil later so you hang in the air a bit longer. interesting to see which moves you can fit into a jump, if doc had SH bair -> uair he'd wreck ppl

 

[schmooblidon]

good to know that how fast you do your cape when SHing will alter how long you stay in the air. I'm not sure what in-game applications this has though

If i'm understanding all this correctly, if you cape asap in a SH you probably won't be able to fit in any aerials after the cape, but you'd still be able to waveland, which might be useful for being speedy when gaining distance (vs falco laser or something) whereas if you're caping and want to follow up with some aerial, you want to start the cape a lil later so you hang in the air a bit longer. interesting to see which moves you can fit into a jump, if doc had SH bair -> uair he'd wreck ppl

Actually you can just about squeeze in a Nair hitbox, when caping on the first frame. You will be able to nair when caping on airborne frame 1-14. Unfortunately it is impossible to upair after a cape, it's ecb pattern is a gift and a curse.

On another note, very strangely... When you do sh cape>aerial, the next cape you do won't rise until you have been in the Landing state (Impact land i.e. empty hop, autocancel), LandingFallSpecial state (airdodge/wavedash/up-b) or you land during a cape animation. It seems like those are the three things that trigger to rising cape refresh. That must be a development oversight.

Edit: Haha this is pretty goofy, you can keep it as a non-rising cape by fullhop cape > aerial/pill repeatedly. If you could wavedash it might of actually had a use.

 

[schmooblidon]

Today magus blessed us with the ultimate frame data mod (see here) so now I can give extremely accurate values for literally anything. I thought I'd test doc's speeds with different movement methods and see exactly how they compare.

Walk

Walk is a pretty intricate mechanic and I have yet to figure out it's equation. Like dash and run it is directly affected by your stick's position and will increase/decrease your velocity accordingly. It decreases speed very linearly and will drop the velocity by 0.06 until the velocity matches the input. The rate of increase however, exponentially decays (this means it gets harder to speed up the faster you go). The velocity caps at 1.10000 but it takes about 3/4 of FinalDestination to reach that speed, but only about 1/8 to reach a velocity of 1.0000.

Dash

There are 17 horizontal input values on each side of the control stick that initiate a dash. But you only start moving on frame 2, so the first input does not affect your initial velocity (your momentum on the previous frame is repeated). Frame 2 velocity onward are dependent on where your control stick is.
When you smash the control stick you will likely either be inputting the horizontal value 1.00000 or 0.98750 depending on your stick, they will give you a velocity of 1.50000 and 1.48125 respectively. Whereever the stick is on frame 2, you will start with an initial velocity of 1.44000, and from then will decrease/increase depending on where the stick is. There is some quite complicated equations for this, but I've save them for another time.

Foxtrot

A perfect foxtrot would be performed by dashing and holding the stick forward for 10 frames, releasing it to neutral, then full forward again on frame 22 to initiate another dash. Assuming you hit the perfect horizontal value 1.00000, your velocity will be 1.50000 until frame 11. From there it will drop by 0.06 each frame and then on frame 22, will repeat the previous frame's velocity of 0.08400 before jumping back up to 1.50000 the next frame. This gives an average velocity of 1.28000.

Run

Run uses the same values as dash and generally works the same way where you decrease/increase speed depending how far forward your stick is, but it does this at a much slower rate. I haven't figured out exactly how it calculates the velocity when changing inputs, it seems pretty complicated, but it does not matter for this. You will want to keep the stick all the way forward and maintain a velocity of 1.50000.

Wavedash

Each frame when velocity is above 1.1, your velocity will drop by 0.12, when it is below 1.1, it will drop by 0.06.

With a perfectly angled grounded wavedash, your initial velocity is 2.67039 and it drops by 0.12 each frame. So by the time you could wavedash again, frame 14, you will have a velocity of 1.11039. Doing perfect wavedashes repeatedly would give you an average velocity of 1.89039.

Using the diagonal notch for a wavedash, your initial velocity is 1.90116 and drops by 0.12 until it reachs 1.06116 on frame 8 and drops by 0.06, by frame 14, your velocity is 0.70116. This gives it an average velocity of 1.21116.

Perfect Waveland

A perfect waveland is when you can hit left/right with an airdodge to waveland and requires very precise positioning, either from a backwards jump or onto a platform.

Initial velocity is 2.79000 and drops by 0.12 where by frame 14, velocity is 1.23000. Giving an average velocity of 2.01000.

Waveland with angles will give you the same speeds as from the ground but for every frame you are airdodging in the air you lose pv/10 where pv = previous frame's velocity. This means generally you will lose speed much faster then when on the ground.

Buffer Rolling

Oddly, the total velocity across all 33 frames adds up to 33.00000 even though each frame has an obscure decimal. So the average velocity is 1.00000.

Moonwalk

Just for fun, a perfect moonwalk would be performed by a perfect waveland backwards, dashing forward on the first actionable frame then on the second frame of dash hitting full back and holding until frame 22 of dash, and repeat to chain them. Frame 1 will take the velocity of the previous frame which in this case is 1.71000, the following frame will have a velocity of 0.29000 and will increase by 0.08 each frame until frame 18 where it caps at 1.50000. The average velocity of the first moonwalk is 1.04524.

Chained moonwalks will follow the same pattern except frame 1 repeats the previous frame's velocity, so it starts at a 0.21 slower velocity. The average velocity of chained moonwalks is 0.86571.

Conclusion

From slowest to fastest

Chained moonwalks......... 0.86571
Buffer rolling............ 1.00000
Initial moonwalk.......... 1.04524
Max Walk Speed............ 1.10000
Diagonal notch Wavedash... 1.21116
Foxtrot................... 1.28000
= Run..................... 1.50000
= Dash.................... 1.50000
Max angle Wavedash........ 1.89039
Perfect Waveland.......... 2.01000

I'm also going to add a new section to the guide for random little stuff which I will update whenever I find something interesting that is either very small or has no beneficial application. I've added a bunch now that you may or may not of known.

When I finally get new hardware to make gifs I will likely drastically change the overall structure and concept of the guide, and make it more about applications rather then advanced tech frame data, like kadano's marth guide. It will likely find a balance between his sort of thing and the way it is now.

 

[#HBC | Mac]

tl;dr: everyone master your max-angle wavedashes

 

[Diana's Safe Landing]

Amazing thread! I have a question on Marios Up-b walljump.

"When up-b is NOT started against the wall
DI Full Horizontal
ECB must begin touching wall on frame 23 (second to last hitbox frame)
Can then walljump on frames 24-29"

So does this mean that when its not started on the wall the only way to wall jump is with the full horizontal up-b? Also since the ECB has to start touching the wall on a specific frame that means that the spacing with the up-b is the absolutely critical part? What counts as truly touching the wall on start up? I feel like on fountain I can start the up-b "touching" the wall but I cant walljump it with the same timing on yoshis. Is there a list of places where the "starting on wall" wall jump works? Does the "not started" on wall one work everywhere? Sorry for so many questions lol but this tech is pretty hard and my understanding of it is vague.

 

[schmooblidon]

Amazing thread! I have a question on Marios Up-b walljump.

"When up-b is NOT started against the wall
DI Full Horizontal
ECB must begin touching wall on frame 23 (second to last hitbox frame)
Can then walljump on frames 24-29"

So does this mean that when its not started on the wall the only way to wall jump is with the full horizontal up-b? Also since the ECB has to start touching the wall on a specific frame that means that the spacing with the up-b is the absolutely critical part? What counts as truly touching the wall on start up? I feel like on fountain I can start the up-b "touching" the wall but I cant walljump it with the same timing on yoshis. Is there a list of places where the "starting on wall" wall jump works? Does the "not started" on wall one work everywhere? Sorry for so many questions lol but this tech is pretty hard and my understanding of it is vague.

It would seem like there are more setups, but I have tested this quite thoroughly and have yet to find them though. This is by far one of the hardest things I've ever tested, it took me a really long time to figure out the frames on the horizontal DI setup.

Spacing is stupidly precise, I can't even imagine doing this consistently.

Your ECB must be literally touching the wall, this typically has your model partially clipped through.

This something I need to test more, but I think the rule is start against the wall, the middle doesnt matter, then before you walljump your ECB must be touching again.

As long as there is a wall then yes it will work with the right spacing. Theoretically you could climb really high walls or stall for a long time under a stage.

 

[Diana's Safe Landing]

I guess the reason I can't do the "started while touching wall" jump on the top of fountain is -because that curved part likely makes it so Mario's ECB isn't "touching" enough. Thanks for the info!

 

[Diana's Safe Landing]

Anyone willing to share the exact percents uthrow uair works on Mario?

 

[#HBC | Mac]

from this reddit thread: http://www.reddit.com/r/SSBM/commen...s_becoming_more_and_more_discouraging/cozm238

As an example look at PPU's pivot F-smash play against puff. Doc's d-throw -> pivot f-smash can be used against any mid-weight character easily, but no doc does that.

how true is this?

more specifically I guess, are there any percentages/character combinations where doing dthrow -> fsmash could likely net a KO (or solid edgeguard opportunity)?

I usually just go for the combo if i don't think dthrow -> fair will kill -- but if there are percents where dthrow->fsmash could kill at lower percents than fair, id def integrate that into my game

____
edit:
also during up-b cancel, you're considered to be in jump squat / crouch (or w/e it's called).. does that mean if you're hit during the upB cancel and you're holding down that you'll get a full crouch cancel?

 

[DocDK]

Great thread man, so so helpful

 

[tauKhan]

also during up-b cancel, you're considered to be in jump squat / crouch (or w/e it's called).. does that mean if you're hit during the upB cancel and you're holding down that you'll get a full crouch cancel?

1. Kneebend (jumpsquat) doesn't have crouch properties
2. As soon as you up b, you're in the up b animation which definitely has no crouch properties
3. You only need to go into Kneebend if you upb OoS

 

[schmooblidon]

how true is this?

more specifically I guess, are there any percentages/character combinations where doing dthrow -> fsmash could likely net a KO (or solid edgeguard opportunity)?

I usually just go for the combo if i don't think dthrow -> fair will kill -- but if there are percents where dthrow->fsmash could kill at lower percents than fair, id def integrate that into my game

I have a feeling it will only work against chars that can be dthrow chaingrabbed. I tested sheik for now.

DI Behind

Damage% before throw | Frame window
40-45 | 1
46-51* | 0
52-63 | 3
64-100+ | 4

*possibility of phantom

For DI forward, you can just add 1 to all the frame windows.

52-63 window sounds pretty legit, could be used out of a cg for a kill or easy edgeguard.

Oh **** I forgot about the port priority. These percents are with a lower port, which means you could add 1 to all the frame windows if you have a higher port.

edit: and i also forgot about staling. which i also need to test for cgs. thats a lot of work, but I'll do that soon

 

[#HBC | Mac]

64-100+ seems really good for sheik. though you're saying that's 4 frames input the pivot fsmash... i'm not sure if that's enough time to consistently react with the correct pivot fsmash input

 

[schmooblidon]

64-100+ seems really good for sheik. though you're saying that's 4 frames input the pivot fsmash... i'm not sure if that's enough time to consistently react with the correct pivot fsmash input

You best fight for that port priority then

Also I needed to fix those movement values

Doc/Mario

Average speeds:
Chained moonwalks......... 0.86571
Roll backward............. 1.00000
Roll forward.............. 1.00004
Initial moonwalk.......... 1.04524
Max Walk Speed............ 1.10000
Diagonal notch Wavedash... 1.25712
Foxtrot................... 1.28000
= Run..................... 1.50000
= Dash.................... 1.50000
Dash>Wavedash............. 1.77384
Max angle Wavedash........ 1.89039
Perfect Waveland.......... 2.25000

Dashdancing:
Dash dance length: 13.665
Average dash dance velocity: 1.24227
Dash frames before run: 10

Extended Dash dance (total 20 frames):
Extended Dash dance length: 25.44
Average extended dash dance velocity: 1.21143

Dashing:
Initial Dash velocity: 1.50
Dash acceleration: +0.08 per frame
Maximum Dash velocity: 1.50

Wavedashing:
Velocity when High Traction is in affect: >1.10
High traction: -0.12 per frame
Low traction: -0.06 per frame

Perfect waveland total distance: 37.92
Max angle wavedash total distance: 35.14209
Diagonal notch wavedash total distance: 21.48075

If you wonna see how he compares to everyone else, check out my analysis, just finished today.

 

[CAUP]

Well I don't know where to post this. I was playing random character, all random stages, items on with my brother, messing around. I was Zelda and he was Mario.I had just gotten a big mushroom. We were on Hyrule Temple, near the slopes leading down to the cave of life, he was below me. He used up b and it sure looked like he up b cancelled for all practical purposes. It happened in a split second but we both saw it and were confused. I've tried to replicate it but I haven't been able to.

If anyone has any idea how this happened I would love an explanation. I'm guessing it has something to do with the slope and the weight of giant Zelda? It was a cool experience. There is a chance we both saw the wrong thing. He did an up b but maybe he was hit immediately out of it? It was confusing. But he definitely did his up b and went no where.

 

[schmooblidon]

I dno dude. I can't think of a reason why it would work. I don't even know why doc's works honestly.

 

[Maple/DBK]

goddamn that haxdash though, makes me jealous to not be able to do it in handwarmers.

 

[1MachGO]

What is Doc's theoretical answer to Fox/Falco platform camping?

Edit: and does he have any moves that reliably out prioritizes their bairs?

 

[Diana's Safe Landing]

Picking pokemon or FD can really help. Uair is like the only viable move to challenge them on the top platform and even that isnt an amazing option. I think dair can beat out bair if you don't let them space around you but honestly I feel like bair beats all of marios moves for the most part. Uptilt is decent at trading with it.

edit: realized you said doc not mario but this should still apply..

 

[#HBC | Mac]

yea platform camping is really hard for doc to deal with. like diana said, stages without a top platform are good. also small stages where it's easier to reach the top platform like yoshis and fod are also good. avoid dreamland at all costs

maintaining center stage control is pretty important vs platform campers. you want to be able to threaten all the platforms from where you are with your aerials and controlling the center of the stage allows you to do that.

so you can either attack from under the platform with an aerial are you can waveland onto the platform and combatting the opponent from there. simply jumping around in the center of the stage allows you to leverage all these options and attack the platforms if you choose to. Note that when you go up onto a platform you're generally the one at a disadvantage and attacking from under the platform is safer. Though it's more likely that you'll be able to punish if you go directly onto the platform though.

generally what I want to do vs spacies is make them jump on the platform and then punish them where they land. to do this, i try empty jumping / double jumping near center stage so they jump away thinking i'm going to hit them. once they jump you can try to predict / react to where they're jumping to and attack that space.

also throwing pills with a trajectory that will cover side platforms is super important. if you're getting the spacy to jump you want to try to limit the options to where they can jump to and placing a pill on a side platform will make them think twice about jumping there so it should be easier to corner them in the air.

these are just some things i try to do, generally platform camping destroys doc and even if you do these things i mentioned a good platform camper will still be very difficult to deal with.

 

[1MachGO]

Yeah, its a pretty **** situation for Doc.

Would love to see a breakdown of some of his hitboxes vs. Fox/Falco bair. I do wonder if Doc's bair or dair can beat it with their disjoint

 

[schmooblidon]

Long awaited comparison images

http://i.imgur.com/MK8TZ9D.jpg

http://i.imgur.com/Ksqv606.jpg

 

[Diana's Safe Landing]

Does anyone use dash grab instead of JC grab in general? Also how does boost grab compare?