Dash to Anything Frames

[Fox Is Openly Deceptive]

A friend (@Teatra) asked me how many frames it takes each character to be able to dash back in the opposite direction after initiating a dash (without using an extended dash dance and of course without getting a dash turn). This got me interested in finding the best possible way for each character to do this and while I was at it, the data would also tell me the fastest possible way for each character to be able to use any action at all after they have inputted a dash (excluding a PP).
(For reference, to do a frame perfect extended dash dance, Cloud would need to re-dash on frame 16 then input the dash back on frame 17, which is 1 frame faster than what he can do below, but in practice? good luck doing that.)

I have the data for 4 different methods below.
The first is inputting a dash, immediately releasing the joystick to neutral and waiting for the skid.
The second is inputting a dash, holding the joystick forwards and returning it to neutral on the frame after you transition into a run then waiting out the skid.
The third is inputting a dash, holding the joystick forwards and inputting shield so you shield on the dash to run transition frame, then immediately releasing shield and waiting out the shield drop frames.
And the fourth is inputting a dash, quickly (on frame 4 of the dash in my tests) move the joystick to the diagonal down/forward indent and holding it there, returning the joystick to neutral on the frame after you transition into a run then waiting out the skid.
The first method I quickly labbed out myself. The second method I looked at the dash length of each character http://kuroganehammer.com/Smash4/DashLength and at first I was just adding the running skid length onto it https://smashboards.com/threads/skid-stop-to-shield-up-times.436583/#post-21190367, but then I found that this didn't always give the right result (see Jiggs) so I manually labbed everyone. For the third method I looked at the dash length of each character and then added 18 frames onto it, which is the minimum shield hold time. Finally for the fourth method I labbed it out manually; note that there was an issue with consistency which I go into later.

The numbers below are the frames on which you can dash backwards (or do anything at all like Jab, U-tilt and D-tilt).


Method 1:
(inputting a dash, immediately releasing the joystick to neutral and waiting out the skid)

Spoiler

You should use this all the time
18 - (limit and non-limit)
20 -

Still worth using this for movement, while being great for attacks
22 -

Esam
23 -

There are better things you can be doing for movement, but it's still relevant for attacks
24 -
26 -
27 -

Not really worth it any more
28 -
29 -
30 -
31 -
32 -
34 - (all MArts)
38 -
39 -
40 -

Lol
45 -

Method 2:
(inputting a dash, holding the joystick, releasing it as soon as you transition into a run then waiting out the skid)

Spoiler

You should use this all the time
18 -

Still worth using this for movement, while being great for attacks
22 -

There are better things you can be doing for movement, but it's still relevant for attacks
24 -
26 -
27 -

Not really worth it any more
28 -
29 -
30 -
31 -
32 -
33 -
34 -
35 -
36 -
37 -
38 -
39 -

Lol
49 -

Method 3:
(inputting a dash, holding the joystick, holding shield to shield as soon as you transition into a run then immediately dropping shield)

Spoiler

There are better things you can be doing for movement, but it's still relevant for attacks
26 -
27 -

Not really worth it any more
29 -
30 -
31 -
32 -
33 -
34 -
35 -

Method 4:
(input a dash, quickly (on frame 4 of the dash in my tests) move the joystick to the diagonal down/forward indent and hold it there, then return the joystick to neutral on the frame after you transition into a slow run.)

Note that this method is inconsistent due to joysticks having a difficult time making consistent diagonal inputs and the corresponding frames of the skid being very dependant on how fast you were running which is connected to the very specific X values of the joystick, and so the results may vary by a frame or two; however, I have accounted for this by testing each one multiple times to get the best estimate and I have also been lenient on the frame I moved the joystick to the diagonal, so you may well get a slightly faster one.

Spoiler

You should use this all the time
16 -

Still worth using this for movement, while being great for attacks
22 -
23 -

There are better things you can be doing for movement, but it's still relevant for attacks
24 -
25 -
26 -
27 -

Not really worth it any more
28 -
29 -
30 -
31 -
32 -
34 -
36 -
39 -

Lol
49 -

Method 4.2:
(input a dash, immediately (on frame 2 of the dash in my tests) move the joystick to the diagonal down/forward indent and hold it there, then return the joystick to neutral on the frame after you transition into a slow run.)

See method 4 above. It's the same thing but stricter timing, and the result is that some characters get a slightly better version.

Spoiler

You should use this all the time
16 -
20 -
21 -

Still worth using this for movement, while being great for attacks
22 -
23 -

There are better things you can be doing for movement, but it's still relevant for attacks
24 -
25 -
26 -
27 -

Not really worth it any more
28 -
29 -
30 -
31 -
32 -
34 -
36 -
38 -
42 -

Final Results using the best method for each character:

The number of the method used will appear beside the character icon if it is either method 2, 3, 4 or 4.2.
If method 4 is equal best with another method, I will not list 4 as it is more difficult to use. For the same reason, if 4.2 is the same as 4, I will not list 4.2. You should still check all your character's methods and decide which one you would like to use. For example, there are a number of cases where method 4 is only one frame faster than a much easier method, so you might decide that it isn't worth it for you.

You should use this all the time
16 - (4)
18 -
20 -

Still worth using this for movement, while being great for attacks
22 - (4) (2)
23 - (4) (4)

There are better things you can be doing for movement, but it's still relevant for attacks
24 - (2) (2) (4.2) (4) (4) (4) (4)
26 - (4.2)
27 - (3) (4) (4.2)

Not really worth it any more
28 - (4) (4) (4)
29 - (4) (4)
30 -

 

[MarioManTAW]

Any idea where the Mii Fighters stand on this? I know when I play in particular that the skid lasts a lot longer than you'd think.

 

[Peabnut Bubber]

how fast can you speed up method 2 if you hold downwards on the control stick to get out of the skid animation faster?

 

[Fox Is Openly Deceptive]

Any idea where the Mii Fighters stand on this? I know when I play in particular that the skid lasts a lot longer than you'd think.

I've added the miis (I used the default size). You'll be pleased to know that method 2 will cut down sword's skid time considerably.

how fast can you speed up method 2 if you hold downwards on the control stick to get out of the skid animation faster?

Holding down doesn't affect this.

 

[Locke 06]

Holding down doesn't affect this.

He's referring to slowing down the run. The thread you linked to in "skid stop to shield times" has a post on it. The old OneSmash video that had a demonstration of it is not available (due to copyright issues probably), but I distinctly remember them using Luigi as an example.

 

[Fox Is Openly Deceptive]

He's referring to slowing down the run. The thread you linked to in "skid stop to shield times" has a post on it. The old OneSmash video that had a demonstration of it is not available (due to copyright issues probably), but I distinctly remember them using Luigi as an example.

Oh I see now. I hadn't thought of that because I was only technically running for one frame in all my test examples. But I wonder if it's possible to make that one frame a slower run..
Yes. Yes it is.
It matters how quickly you put the joystick into the diagonal following the dash, which isn't an issue as I can guarantee the same inputs on the same frames every time, and it matters how far you hold towards the forwards input which can be overcome to some extent by moving to the diagonal down indent. This stuff matters in-game though imo as the difference in a few frames will mean the difference between e.g. a jab or a dash attack, so you want these timings to be consistent.

But here's where I'm running into a problem. I've been getting inconsistent results if I input the diagonal on the second or third frames of the dash, even though I'm making the exact same inputs on the exact same frames every time (I even tried restarting after each test). I obviously didn't have this difficulty getting consistent results earlier, and the interesting part is I don't have any trouble getting consistent results when inputting the diagonal on frame 4 of the dash or later.

Having said that, with Cloud for example, his method 2 gave him 35 frames, but with this (method 4), by moving the joystick into the diagonal on the second or third frame of the dash, I was able to get anywhere between 28 and 30 frames for both, and if instead I moved the joystick diagonally down on the fourth frame of the dash I was able to get frame 29 every time [though now I have to wonder if I was just lucky].

I'll see what I get for everyone else, and assuming this bizarre result is the same for other characters, I might end up getting the results of method 4 by inputting the diagonal on frame 4 for everyone and be done with it.
Edit: nope. Bayo is getting inconsistent results from frame 4 too. I really don't understand. And I really need to draw the line somewhere otherwise this will never get done. So I think I'll just input on frame 4 and do 3 tests for each character and then list the highest number I got. It'll only be off by a frame or two at worst depending on what I can only describe at this point as RNG.


(Ftr I was unable to make Sonic's method any faster no matter what I did, so perhaps this doesn't affect everyone? perhaps his was already optimised? we'll see.)

Edit 2: Arthur reckons it's the minute differences caused by joysticks being terrible. Not much I can do about that.

 

[Fox Is Openly Deceptive]

Updated with method 4 and a few corrections. There have been some significant changes and the final results are looking much better, making some form of dash movement viable for a good portion of the cast, much to my amazement.
Eventually I'll probably store this in the academy for reference, but let me know at any point if there's anything else that needs to be done.

 

[Locke 06]

Luigi doesn't look right for method 4.

 

[Fox Is Openly Deceptive]

Luigi doesn't look right for method 4.

I know what you mean, I was surprised too so I checked it multiple times and always got 49. It's possible that it might just have something to do with starting the skid on the frame after the transition to a run as opposed to waiting a bit to possibly slow down more. I'll look into this quickly.

I'm still getting 49 when I do it the way I outlined for method 4, i.e. dashing for 3 frames then moving it into the diagonal indent until the dash to run transition on frame 11 then returning it to neutral. And then when I checked what happens if I keep it in the diagonal until frame 12 or frame 15, I actually still got frame 49 for when the skid stopped, so while I definitely decreased the skid time, overall (counting from the moment I dashed) I didn't save any frames coincidentally.
Having said that I'll bet that it might make a difference if I dashed forward initially for one frame only then moved it down to the diagonal until frame 11, just because he does a lot of his accelerating in the first few frames. Yeah there we go, frame 41, 43, 44, 44, and here we have all that inconsistency I was hoping to avoid which is now being accentuated.
Hmm... I wonder if it would nonetheless be worth getting a 'method 4.2' or something, which would essentially just be an average of the results I get for inputting the dash forwards for one frame only.. It probably would be worth it. Well alright, I guess I'll get onto that when I can.