Hitstun/shield stun Thread

[MuBa]

So, how's the code holding up so far?

Hitstun hasn't been added BUT there has been hitstun on all types of animation when being hit.

Although I see Heavy Brawl being fantastic for this code, but it still needs a bit more tweaking.

It's definitely an improvement from the previous version.

 

[kupo15]

Don't forget that a sweetspotted Utilt is much stronger than a sour one.

I tested this again.
Fresh Ike Fsmash on Snake at 20%- approx 1.11 secs of recovery
Completely stale Fsmash on Snake at 20%- approx 0.67 secs of recovery

With some error in my windows movie maker, conclusion is that the hitstun multiplier is from 0.5-1.0
This means that we probably could get a hitstun modifier and be set!!

 

[Revven]

With some error in my windows movie maker, conclusion is that the hitstun multiplier is from 0.5-1.0
This means that we probably could get a hitstun modifier and be set!!

This would be epic, I've wanted a hitstun modifier for the longest and if it is multiplied, then we'd have floating points which would = awesomeness.

 

[sagemoon]

Still too little hitstun imo.

 

[Starscream]

With standard floaty Brawl, not enough hitstun. With more gravity it's better.

 

[leafgreen386]

Still too little hitstun imo.

Um... because this code isn't meant to actually increase hitstun, just to make it so you can't airdodge or aerial while you're currently in hitstun. We should be getting another code soon that we will be able to actually modify the amount of hitstun, so that's something to look forward to.

 

[MBlaze]

New code is beautiful, just need the modifier and thats it for hitstun.

 

[Magus420]

I have an idea for more stun time. Im pretty sure the stun time gets less the more stale the move is so what should work is if the stun time is equal to that of a fresh move all the time no matter how stale the move is.

I tested this again.
Fresh Ike Fsmash on Snake at 20%- approx 1.11 secs of recovery
Completely stale Fsmash on Snake at 20%- approx 0.67 secs of recovery

With some error in my windows movie maker, conclusion is that the hitstun multiplier is from 0.5-1.0
This means that we probably could get a hitstun modifier and be set!!

Sure the stun just isn't based directly on knockback power? A stale move hitting at the same damage will obviously have less knockback, and I would assume less stun as well. I don't really see where you're going with the stale moves thing. To have it have the same stun as an unstale hit of the move you'd probably have to go about it by making the knockback the same which would then be used to calculate hitstun, which wouldn't really do anything besides remove the stale moves effect. Just adjust the way stun is calculated, and add a small multiplier or something.

 

[kupo15]

well, no, we are not sure if knockback is related to hitstun in coding. Here is the PM I sent PW.

So, can you make a code that allows us to freely modify the value controlling the stale moves variable in the hitstun calculation and that will always override what the game would normally use for that variable? Please make sure that it strictly affects hitstun, and not stale moves as a whole. We don't want the damage or the knockback portion of the stale move system to be affect in any way.

He said that it may be possible

 

[leafgreen386]

It would be bad if it was based directly on kb, however I'm going to assume (read: hope) that they kept the hitstun calculation system from melee which had kb and hitstun as two separate components.

 

[Magus420]

How are you sure there even is some kind of stale moves variable for calculating hitstun? That test you did doesn't really show anything. Instead of comparing the stun of an unstale hit and a stale one with 2 completely different knockbacks, increase the damage they have for the stale one so that both knockbacks are equal or nearly equal. If they are noticeably different yet have nearly the same knockback you'll know the fact that it's stale has an effect on the stun itself.

 

[MuBa]

So basically you're asking that we should make a code that won't affect moves being staled? Good Lord, Snake would be uber then O_O

 

[kupo15]

How are you sure there even is some kind of stale moves variable for calculating hitstun? That test you did doesn't really show anything. Instead of comparing the stun of an unstale hit and a stale one with 2 completely different knockbacks, increase the damage they have for the stale one so that both knockbacks are equal or nearly equal. If they are noticeably different yet have nearly the same knockback you'll know the fact that it's stale has an effect on the stun itself.

Wow Magus, you have excellent points. You know your stuff. I tested it and there is really no change....

Fresh Fsmash at 20% had the same distance and stun time as a staled Fsmash at 70%

So what does this mean?

 

[Magus420]

Well, I'm fairly positive the hitstun formula in Melee is something as simple as...

Hitstun = Launch Speed/1.475

...and rounded up to the next whole frame.

Also, for attacks that bounce you off the floor the actual launch speed is reduced by 20%, but calculates the stun off the original launch speed, so with those moves you end up with 25% more stun than normal for the same knockback power.


You can find out that number by the "Top Speed" (person hit) or "Fastest Pitch" (attacker) stats on the results screen assuming the knockback in question was the strongest one in the match. Dunno if there's something similar to that in Brawl.

If there is something like that and you can find a relation between it and the number of frames the knockbacks stun you (the real hitstun not the cancelable crap), then you should be good to go.


If it's still similar to that you could just change that number to adjust the hitstun to knockback ratio to whatever you like.

If that's possible you'd really want to somehow include some coding to have it only apply to when the launch speed is above the amount that puts characters into a tumble/makes knockbacks DIable, that way you wouldn't be adding more hitstun to stuff like Falco's/Pika's CGs, Sheik's f-tilt lock, and other dumb combos you can't really DI effectively.

If down the line a code can be made to make those kind of weak hits DIable like normal then you could remove that part of the coding and have the hitstun change apply to all knockbacks safely.

 

[leafgreen386]

Wow. Is that really it? We better inform pw, then, that hitstun is more than likely not calculated in the way we thought it was. But that's excellent to know, magus. It likely has stayed the same (except maybe for the 1/1.475 part).

 

[Magus420]

Yeah, it really is that simple apparently, lol. Between that stun formula I figured out and this...

http://www.smashboards.com/showthread.php?t=99071

...you can actually calculate the stun time on moves in any hypothetical situation (assuming all of those numbers are correct of course). The ones I tried out matched up perfectly from start to stun time

 

[kupo15]

Magus' awesome epic post

Wow, I feel dumb for suggesting the stale move system...too good (like Isai good)....your the man Magus! Remind me to shake your hand if I see you around NJ

And yes, Top Speed is still in the game

Also, for attacks that bounce you off the floor the actual launch speed is reduced by 20%, but calculates the stun off the original launch speed, so with those moves you end up with 25% more stun than normal for the same knockback power.

Did you mean wall?

 

[Magus420]

Nah I meant floor. You know, like Falcon/Ganon's stomps and similar moves. If they're grounded, they bounce you off the floor and send upwards.

 

[kupo15]

ooo, in that situation.. got it. How does the wall work? This code should come out really fast. Hackers make modifier codes in no time and you spelled it out for him how to find the value.

 

[Magus420]

Bouncing off of walls also reduces your flight speed by 20% but keeps the same stun. Ceilings probably work the same way as well.

Actually... looking at this I imagine a lot of this stuff probably works the same in Brawl as in Melee. I knew I also saw the 20% thing around here somewhere.

Grounded Spikes

When you do a spike on a grounded foe they are popped upwards. The knockback that they travel upwards is actually less than the knockback of an aerial spike. 20% less to be more precise. This seems to be the trend with all the spikes I have tested (as seen below), but I am not finished yet.

 

[kupo15]

This is a well deserved *bump*

EDIT: I tried to figure out what the hitstun constant in 64 was but that game has no "Top Speed" results

 

[Onomanic]

Any videos yet? XD

 

[kupo15]

Any videos yet? XD

Its not much different than 1.2. The gameplay is the same

Magus: I am trying to find the hitstun constant for brawl but Im having trouble.

Ganon warlock punch snake at 0%. Max launch speed is 4097 mph, stun time is .87 secs or 52 frames.

C=launch/hitstun time

Am I missing something?

Yeah, it really is that simple apparently, lol. Between that stun formula I figured out and this...

http://www.smashboards.com/showthread.php?t=99071

...you can actually calculate the stun time on moves in any hypothetical situation (assuming all of those numbers are correct of course). The ones I tried out matched up perfectly from start to stun time

Does that mean you know what the hitstun constant is or can figure it out?

 

[Magus420]

I know how it's calculated in Melee. I don't have Brawl to be able to figure that out.

 

[kupo15]

hmmm, I wonder if this "fastest pitch" in melee and "top speed" are the same. They have to be but the numbers are so much higher.

Marth Fsmash(tippered) against Pit who is in the middle of the weight list
0%: 3695 mph (top speed)
Damage 19%

In melee, the Pitch speed was 65 I believe

 

[Magus420]

Fastest Pitch is how fast you sent someone else, and Top Speed is how fast you yourself were sent. The fact that the actual numbers were far lower in Melee doesn't really matter since the units are made up and only relevant within their respective games.

How long was Pit stunned from that f-smash? Was it anywhere around 47 ish?

Also, it's important that you don't SDI. Dunno about Brawl, but Melee counted SDI as a speed. I noticed it once and was like 160ish which was faster than the attacks I was testing with.

 

[kupo15]

here is our data:
Fresh tipper on pit

Dam(post hit) ............................................ Stun time......................... Top Speed

19%....................................................23 frames.......................3965
40% .................................................... 27 frames ........................ 4418
70% .................................................... 35 frames ........................ 5501
109% .................................................... 44 frames ........................ 6892
320% .................................................... N/A........................ 14558

I did not do any DIing

 

[leafgreen386]

No. It was 23 frames, actually.

Since the hitstun formula is
h = l/c

Where h = hitstun, l = launch speed, and c = a constant

Then l = ch

So in plotting hitstun against launch speed, the c value would be the slope.

Using this data (percent after hit - launch speed - number of frames of hitstun):
19 - 3695 - 23
40 - 4418 - 27
70 - 5501 - 35
109 - 6892 - 44

The closest I've come to a hitstun formula is:
l = 150h + 300

which can be rearranged for hitstun to be:
h = (l - 300)/150

Which is still (probably) imperfect. The last three numbers check out (if and only if the game doesn't use a strict rounding up or down feature), but if this is right, then the 19% launch was actually 24 frames of stun.

edit: I see kupo has already mentioned our data. This is the results of that data. We're currently doing more tests to try to get more accurate numbers.

edit2: More data

(percent - launch speed - frames of hitstun)
55 - 4975 - 30
91 - 6236 - 38

We're probably gonna need at least 10 or 12 points of data before stuff starts making sense.

edit3: A lot more data. All of our data currently:
(damage - launch speed - frames of hitstun)
19 - 3965 - 23
25 - 3904 - 25
35 - 4263 - 26
40 - 4418 - 27
45 - 4615 - 29
51 - 4812 - 30
55 - 4975 - 30
60 - 5139 - 32
65 - 5326 - 33
70 - 5501 - 35
91 - 6236 - 38
109 - 6892 - 44

So with our current data, the model we're getting looks like h = (l - 110)/157

 

[Magus420]

Is this correct though? Seems weird given those other numbers

Ganon warlock punch snake at 0%. Max launch speed is 4097 mph, stun time is .87 secs or 52 frames.

Also, are you guys testing with or without the code?

 

[kupo15]

forget my ganon tests. They were prob wrong. Leaf is updating his post to add two more values.

Something occurred to me out of observation but idk if it has any significance. It would appear that for brawl, they multiplied the top speed and stuff by 100 to get rid of the decimals. Leaf says his current c value is roughly 157. 157/100=1.57...this looks similar to melee. Since the number is higher than melee and this is an inverse graph, this would make sense that brawls c value is higher..

Also, are you guys testing with or without the code?

With hit stun 1.3

 

[leafgreen386]

Ok, I've updated my post with our most recent data. However, I think I'm about ready to say that the c value is actually 165 (or at the very least is between 160 and 165).

Why?

Because I decided to see what the data would come out as if I tossed the higher values (91% and 109%), since there is a high likelihood that there was error in measurement on them. Well, I got
l = 159h + 40.5

Which comes out as
h = (l-40.5)/159

But considering how our intercept keeps getting smaller and smaller the more data we get, I'm quite sure that it isn't supposed to be there at all, and that the c value should be somewhere around 165. I'm about to start figuring out if this matches up with the data we got.

 

[izzy24]

****, it was for stuff like this that motivated Panda to create the "smash lab", u guys should let him know about this experiment and make it the #1 priority there for everyone to help and get better and faster result.

 

[MuBa]

kupo15 and leafgreen, explain me in AIM more about this and I'll help you guys out with the equation for hitstun.

 

[Dan_X]

I can't believe how complicated this stuff is. I have no idea how you get these numbers, count the frames, so on. Keep up the good work guys, bravo on your work thus far!

 

[leafgreen386]

Sorry for the wait. Due to some inconsistencies in measurements, it took longer than we thought it would to get this. We had to redo several of the measurements, of which the corrected values in this list are bolded (with the old value in parenthesis). After this, I'm 99% sure that I was right in my initial suspicion that c = 165. We're in the middle of testing other values right now to make sure this holds true for them, too. For now, here's our data:

(percent - calculated hitstun to one decimal place - actual hitstun)
19 - 3695 - 22.4 - 23
25 - 3904 - 23.7 - 24 (was 25)
35 - 4263 - 25.8 - 26
40 - 4418 - 26.7 - 27
45 - 4615 - 27.97 - 28 (was 29)
51 - 4812 - 29.1 - 30
55 - 4975 - 30.2 - 31 (was 30)
60 - 5139 - 31.1 - 32
65 - 5326 - 32.3 - 33
70 - 5501 - 33.3 - 34 (was 35)
91 - 6236 - 37.8 - 38
109 - 6892 - 41.8 - 42

Assuming the hitstun data is all correct, then each value in this list checks out.

Kupo also did a quick test of bowser's fsmash against tl, and got
(percent - calculated hitstun to one decimal place - actual hitstun)
76 - 5046 - 30.6 - 31

Which also checks out. Although the model that the strict data points produce actually comes out to something like
l = 165h - 125
or
h = (l + 125)/165

None of the numbers we got check out for it, so I think I can say with confidence that the formula for hitstun is

h = l/165

Where h = hitstun and l = launch speed, and the result is rounded up to the nearest whole frame.

Hopefully some other people can test this, too, and verify our results. Kupo should be back shortly to post the method he used to get the hitstun data.

 

[Magus420]

Good stuff!

I'm glad I randomly decided to try and figure it out in Melee the other day and realized the way they do it was so simple, lol.

 

[kupo15]

Good stuff!

I'm glad I randomly decided to try and figure it out in Melee the other day and realized the way they do it was so simple, lol.

Me too! Nice save!

I want to let ppl know what I did on my end in case they want to do their own testing.

1. Decided the starting values you will do, what move and on who
2. Perform the Move while recording it and wait until a couple seconds after the tumble
3. Record the percent after the hit and quit the game
4. On the records screen, find "Top speed" on the char hit and record it
5. In the most sophisticated movie software you have, cut out everything before the character is launched (aka out of hitlag) and cut out everything after (and including) the first frame of the tumble animation. The clip in the software should be in 60 frames/sec but its doable with 30 frames/sec
6. Click on the track and it tells you how many frames it is
7. Give all of this to your mathematician

 

[leafgreen386]

Good stuff!

I'm glad I randomly decided to try and figure it out in Melee the other day and realized the way they do it was so simple, lol.

Thanks. So are we. It made things way easier on us

 

[kupo15]

too good!!

 

[IM_A_HUSTLA]

i dont get what u meant by u had to use the old code because u can still airdodge????

im high, sorry if i just should read it sober