The Smash Lab: What is it?

[Browny]

im gonna have to disagree that stale dthrows have different IASA frames. although i didnt go into detail in the video i posted, i found the IASA frame (or first available buffered frame as i called it) to be the exact same in both cases when comparing about 5 different batches of stale v non stale dthrows

And i believe the differences you are seeing in the animation (as in when they are thrown from DDD) is directly attributed to the angle it sends them

like i said on the general boards, imagine it this way. the Y axis is the characters distance from DDD's body, the X axis is time

the overall distance they slide (line integral) is the same in both cases, but they take different paths. I think the stale version has a notcieable larger initial distance from DDD, however the fresh one quickyl catches up. imagine the frame at which DDD can regrab is about 1/3 long way on the x axis

 

[Kirk]

There IS an extra frame of Hitstun in the Non-Staled Dthrow. You can clearly see this in the video I posted. This is the only reason the IASA frame of the Dthrow is different between Stale and Non-Stale.

It's still technically in the same place...It's just that since there is that extra frame in the Non-Stale Dthrow, the IASA is in turn offset by 1 frame.

Also, edited my previous post.

 

[Browny]

i know the hitstun changes (although it doesnt make a difference in the infinite) but when i tested all my dthrows i found the IASA frame to be the exact same for every type of dthrow, down to 2 frames every second on my editing program

i really dont believe anything of DDD's changes except the angle at which dthrow sends them, its all related to who he is grabbing and their various size/airspeed/weight combinations. which is also obviously the cause of who can can chaingrab in the first place while the angle affects the infinite on mario

 

[Kirk]

I personally wouldn't test in Brawl's slower speed mode, as it's adding in frames (frames that normally wouldn't be there at normal speed) in between to maintain the smoothness of the game, which might skewer your results depending on how you capture it.

And also, in testing, I think it's best to do so that 1 frame is 1/60th of a second (60fps), as that is how fast brawl runs

I do all of my testing at normal brawl speed and capture in 60fps to ensure no mistakes or contradictions.

I just did another test, and it confirms my results. The Non-Stale Dthrow's IASA frame is on frame 47, while the Stale Dthrow's IASA frame is on frame 46. This is due to the 1 frame less hitstun.

 

[Mmac]

From the Video though, it didn't look like that at all (Referring to DJ's Path's). It looked more the Fresh throw path was an Arc (Rising upwards, than dropping), and the Stale look like an inverted version of you're fresh path.

 

[FadedImage]

I'm in agreement with Kirk on this one.

It doesn't really make sense that knockback angle would degenerate. For instance, in what direction would a 90 degree attack degenerate to?

However, reduction in hitstun makes plenty of sense. The lack of hitstun means the character has more time to do their natural slide, as opposed to the hitstun slide.

Take notice: the characters that need to be pummeled have ridiculous ground slides when you hit their shield.

my theory:
lack of hitstun means increased time for characters to slide without inhibitions. when they are in hitstun animation, they slide at a slower rate, meaning they won't go as far.

 

[Browny]

From the Video though, it didn't look like that at all (Referring to DJ's Path's). It looked more the Fresh throw path was an Arc (Rising upwards, than dropping), and the Stale look like an inverted version of you're fresh path.

yeah mmac i quickly realised that lol, my directions were backwards, I fixed it. also the graph isnt a picture of the path they take since this is measuring displacement vs time, not displacement vs displacement.

and kirk I also worded what i said wrongly there too. by 1/50th, i mean i could break down a single frame into roughly 10 frames on my editing program. i dont know how that fraction adds up, but its small

@ faded image, if hitstun is to blame, how come a 999% fresh dthrow has the exact same animation as a 0% dthrow?

 

[Kirk]

and kirk I also worded what i said wrongly there too. by 1/50th, i mean i could break down a single frame into roughly 10 frames on my editing program. i dont know how that fraction adds up, but its small

To me, that seems unnecessary. Splitting frames into smaller pieces doesn't seem like it would help.

Getting more accurate results? I don't see how you can you get more accurate than 1 frame...there isn't anything to look at in between frames (this goes for normal speed brawl, as that is how everyone would be playing it anyhow).

I don't mean to bash your methods, but to me I don't see how it would help.

 

[Vv2]

Stale Moves were put into the game to avoid repetitive spikes and or move sets. as far as it can be determined the stale move "stack" in a list you can clear this list either by dieing or using other moves not by time.the best guess as per how many moves stay in the stack is about 8 or so.

Grab,Punch,Grab,Punch,Grab,Punch,Grab,Punch

1 grab 1 punch =
..." "
..." "
4 grabs 4 punches so stale moves doesn't kick in

Grab,Grab
Consecutive move activate stale moves
Or.
Grab,Grab,Grab,Grab,Grab....punch,kick,over B
...5 grabs the list % has been out weighted with grabs
Run line of code X

 

[Browny]

the aim of that was to remove interpolation between frames. since i know my camcorder is running at a max of 50 fps, i cant tape brawl in full speed since there will be one frame interpolated for every 5. slowing it down as far as i did not only makes the interpolated frames extremely easy to see, you can pinpoint exactly where they begin and end which pretty much removes all error involved. when ive tested my frame capturing methods vs other characters boards they were all accurate to the frame with a max error of about 0.3 frames. and that was only on a single test per attack

with this, i measured the iasa frames 10 times (5 lots of stale v non stale) and they all ended up on the exact same frame (200 or something once i slowed it down). Now if interpolation was a problem, surely in one of my tests there would have been at least 1 interpolated frame within the entire animation (1/5 * 200 = 40) but all mine were the same exactly. If there was at least 1/200 frame difference i might accept my method is not accurate

eh these numbers are too complicated

tl;dr when slowing the game down to 1/20x speed, i found the IASA frame to be the exact same on every type of throw

 

[Kirk]

Ah i see how you do it now. You said you tested it on slow brawl correct? If that is the case, that can cause some error in your results methinks.

In my testing, I just have my set 60fps video, normal speed. Over the past few minutes, I've been looking at other things and doing tests.

I myself, every time, am getting the same results.

Non-Stale:
-5 Frames Hitstun.
-IASA Frame 47.

Stale:
-4 Frames Hitstun.
-IASA Frame 46.

This goes for high damages(300% +) as well, which I just tested...exact same results, in case anyone was interested.

eh these numbers are too complicated

I agree. :/ I still think there is room for error left in that method...but that's just me. Heck, all methods aren't 100% accurate...there's no foolproof way to collecting frame data just yet.

 

[Arturito_Burrito]

I'm the one who told him that hit stun changes in stale moves so I say I become the 2nd smash researcher.

Spoiler

jking great job kirk

 

[Browny]

i give up lol. I really cant argue vs 60 FPS cature methods, no matter how much i slow mine down by. but i still think the trajectory mario takes changes as the move becomes stale. Ill just comprimise, its the combination of both that makes him escape after 5

 

[Kirk]

Alright, I just got done looking at both Mario and Luigi.

Results:

EXACTLY the same. Both Mario and Luigi's results are perfectly consistent with my testing with Samus.

Frame Data:

Non-Stale Dthrow:
-Hit: 21 (5 Frames of Hitstun)
-"P2" Marker Appears: 37
-Dthrow IASA Frame: 47

Stale Dthrow:
-Hit: 21 (4 Frames of Hitstun)
-"P2" Marker Appears: 35
-Dthrow IASA Frame: 46

Note: The "P2" Marker is the "Player 2" icon that usually appears over your character. It disappears as soon as you get grabbed, and reappears when you are "thrown" or released. It's used as an indicator in testing for when the opponent gets "thrown" or released from DDD.

-----------------------------------------------------------------------------

To me, it just looks like they slip out earlier in the stale throw...and are just out of range for DDD to regrab them (misses by 1 frame by the looks of it).

 

[SaltyKracka]

Alright, I just got done looking at both Mario and Luigi.

Results:

EXACTLY the same. Both Mario and Luigi's results are perfectly consistent with my testing with Samus.

Frame Data:

Non-Stale Dthrow:
-Hit: 21 (5 Frames of Hitstun)
-"P2" Marker Appears: 37
-Dthrow IASA Frame: 47

Stale Dthrow:
-Hit: 21 (4 Frames of Hitstun)
-"P2" Marker Appears: 35
-Dthrow IASA Frame: 46

Note: The "P2" Marker is the "Player 2" icon that usually appears over your character. It disappears as soon as you get grabbed, and reappears when you are "thrown" or released. It's used as an indicator in testing for when the opponent gets "thrown" or released from DDD.

-----------------------------------------------------------------------------

To me, it just looks like they slip out earlier in the stale throw...and are just out of range for DDD to regrab them (misses by 1 frame by the looks of it).

Woo, Kirk for the first SR!

 

[thesage]

Didn't people already know that stale moves have less hitstun? >_>;

 

[Browny]

find out if the hitstun frames change on any other character. i mean it does seem kind of odd how bowser cant be infinited while someone as small as mario can, i cbf doing any more testing lost too much sleep already

 

[Kirk]

The answer to the question isn't just the hitstun. It's also when your opponent gets released from DDD.

Did tests with characters that can't be infinited as well as those who can be.

Results:

Surprise, surprise...exactly the same. Read my previous post for the data.

Basically, with the Stale Dthrow, your opponent gets let out earlier in relation to when DDD can act, and therefore your opponent's hurtbox is further from you then it would be for a Non-Stale Dthrow.

For the Stale Dthrow, since your opponent gets "thrown" 2 frames earlier, but DDD can act 1 frame quicker, your opponent nets 1 frame extra time to distance themselves from DDD.

The reason it works on large characters (such as DK), even with the Stale Dthrow, is simply because their hurtbox is very large. Even with the extra frame you get when the Dthrow is stale, it isn't enough to move your hurtbox away from DDD's grab range.

Characters like Mario, Luigi, and Samus can escape when Dthrow is stale because their hurtboxes are small enough to be able to use that extra frame to distance themselves and in turn escape DDD's grab range.

I hope I explained that well enough...

 

[§leepy God]

So this will be another version of the Smash Back Room but talk about the meta game and AT's right? Too much work for me, lol.

 

[comboking]

dude this is a great idea I will get on it right away.

 

[Mmac]

I think Kirk pretty much has it. I still think the Throw Paths have a factor with this also, though.

Sigh, If only I knew how to count frames

 

[K 2]

I would be really interested in doing this, but I don't have the means to figure it out (no capture card, etc). sigh...I'm sure someone's figured this out already...

 

[comboking]

same here man i forgot about that

 

[ShadowLink84]

Its figured out that the opponent being released early or/and the trajectory being influenced by the move staling correct?

Would this also mean the game is assosciaing the same behavior?
How knockback and hitstun are both affected and so the game releases the opponent 1-2 frames earlier from the grab as a method of compensating for the hitstun (since their is a certain amount of time before the opponent can react upon beig thrown).

 

[Ukemi]

So if someone has already won/been chosen, what qualities of a member would you look for to consider them for SL? Or are you going to like hold a contest every month or something?

 

[FadedImage]

I don't have the capture equipment to test frame rates, so it's rather difficult to conclude anything.

However, how does it make sense that the IASA frames get better as the move gets stale? Also, how are you measuring "hitstun"? Has anyone checked whether or not these characters are getting invincibility frames as they're being thrown?

You can infinite fine even with 999% damage and above. So it can be concluded that hitstun isn't based on damage. So is it based on KGR? Since Dedede's d-throw has 0 KGR, it doesn't make sense that hitstun would change.

I don't understand how the game programming would change the IASA frame on an animation due to stale moves, it just doesn't make sense.

 

[Kirk]

-The Infinite works because DDD's Dthrow has set knockback, so nothing would change with a different percentage.

-The stun I am referring to is how long the character is "frozen" in place before resuming back to the normal animation. You can see in the video I uploaded, http://www.youtube.com/watch?v=xaPqbyjiMU8, that DDD is "frozen" for one more frame in the Non-Stale Dthrow.

-Hitstun? Hitlag? No one has corrected the term if it was being misused...even after watching the video. If there is an error in the terminology, I wouldn't mind being corrected. But the "frozen" frames that DDD has during his Dthrow is what I was referring to.

-The IASA of DDD's Dthrow doesn't actually change. It is simply offset by 1 frame because of the different values of hitstun between the Non-Stale and Stale Dthrows. Since there is 1 less frame of hitstun in the Stale Dthrow, the IASA would in turn be offset 1 frame earlier.

I hope that clears things up.

 

[ShadowLink84]

It would be hit lag i think. The point at which the user is still in the frame of the attack even thuogh the opponent has been struck.
We need capture cards XD

 

[FadedImage]

-Hitstun? Hitlag? No one has corrected the term if it was being misused...even after watching the video. If there is an error in the terminology, I wouldn't mind being corrected. But the "frozen" frames that DDD has during his Dthrow is what I was referring to.

ohhhh

I should have watched that vid earlier, lol. yeah, that's actually hitLAG. Sort of... cuz you can't really SDI during it so it's not REAL hitlag, but that's probably the best description.

It's interesting that hitlag would be reduced with stale moves... To prove this, it would be good to test the hitlag frames on another no KGR attack that isn't a throw. I can't think of an attack that has 0 KGR but still high hitlag.

Otherwise, Kirk wins thread. Congrats to Kirk, the first SWF researcher (?). d:

The first person to correctly discover exactly what stales in Dedede's downthrow will become the first accepted member of The Smash Lab!

Kirk's "hitlag".

 

[Adapt]

Yeah I think Kirk has got this in the bag, although i thought I saw a difference in angle during my testing.

I don't have the equipment to measure frame data

Faded: there is none (0 KGR and high hitlag). The simplest no KGR attack is ZSS's single jab. I am doing some testing about character reactions to these types of moves.

 

[Kirk]

Ah thanks for that clarification. I suppose in retrospect hitlag would have been the better term to use. I guess I was just confused on what to call it since it wasn't concerning the person getting attacked. *shrug* But no matter.

I just want to clarify also, that the hitlag isn't the whole answer. The other part is when your opponent gets "thrown" from DDD's Dthrow.

From the end of the hitlag and hitstun from the attack, your opponent starts flashing and jumbling up and down for a bit, eventually moving outwards, until that "P2" indicator appears. The amount of frames it takes to get to this point is less in the Stale Dthrow than the Non-Stale Dthrow (If you watch the video and count the frames yourself, you will see it is indeed faster).

Timed from the first frame of hitlag (to be consistent with both throws):

Non-Stale Dthrow: - 16 frames.

Stale Dthrow - 14 frames.

There's a difference of two frames here. Since the Stale Dthrow has one less frame of hitlag, you subtract that from the two frame difference, and that is where the 1 frame advantage I've been talking about comes from.

Sorry if I seem like I'm repeating myself over and over again, I just want to explain this as fully and clearly as I can.

 

[Mmac]

I have Video Recording and (Good) Editing programs, I just didn't know how to count frames

 

[Browny]

put it in slow-motion (i bring it down to 5-10%) can count them?

 

[Kitamerby]

Ah thanks for that clarification. I suppose in retrospect hitlag would have been the better term to use. I guess I was just confused on what to call it since it wasn't concerning the person getting attacked. *shrug* But no matter.

I just want to clarify also, that the hitlag isn't the whole answer. The other part is when your opponent gets "thrown" from DDD's Dthrow.

From the end of the hitlag and hitstun from the attack, your opponent starts flashing and jumbling up and down for a bit, eventually moving outwards, until that "P2" indicator appears. The amount of frames it takes to get to this point is less in the Stale Dthrow than the Non-Stale Dthrow (If you watch the video and count the frames yourself, you will see it is indeed faster).

Timed from the first frame of hitlag (to be consistent with both throws):

Non-Stale Dthrow: - 16 frames.

Stale Dthrow - 14 frames.

There's a difference of two frames here. Since the Stale Dthrow has one less frame of hitlag, you subtract that from the two frame difference, and that is where the 1 frame advantage I've been talking about comes from.

Sorry if I seem like I'm repeating myself over and over again, I just want to explain this as fully and clearly as I can.

So in layman's terms, DDD is somehow frozen in place for an extra 2 frames, and the opponent gets free one frame sooner, and thus can escape? <<

 

[Browny]

lulz so going over my data is super super ridiculous slow-motion

time between DDD's dthrow damage being dealt and the IASA frames on the first 5 throws were

210
210
210
210
215

So that would explain the 1 frame difference kirk found

...
I want a better camcorder

(I still think they follow a different path on release)

 

[Mmac]

Could be the More Damage = More Freezeframes properties. As with heavy damaging attacks, If you get hit by one (For Example Ike's Fsmash), your character freezes for a few, then get launched, where as weaker attacks react almost instantly. Could be the same thing here

 

[Kirk]

In layman's terms, treat the hitlag as the offset (because that's what it is inherently doing, offsetting the Dthrow animation by 1 frame) and know that a Stale Dthrow "throws" your opponent 1 frame faster.

In SUPER layman's terms, with a Stale Dthrow, your opponent gets released one frame quicker, thus providing more distance from DDD for escape.

Now I'm off to bed...thanks for all the support and I hope I explained everything clearly enough. :D

 

[M.K]

Could be the More Damage = More Freezeframes properties. As with heavy damaging attacks, If you get hit by one (For Example Ike's Fsmash), your character freezes for a few, then get launched, where as weaker attacks react almost instantly. Could be the same thing here

You could think of an attack like lighting a fuse to a cannon.
When you "light the fuse of the cannon" with Ike's FSmash at Lower percentages, the cannon fires quickly and not as far.
When you "light the fuse of the cannon" with Ike's FSmash at Higher percentages, there is quite a significant BURST of power and the recoil produces a "slow-motion" effect.

 

[Helkulkhamen]

I really like the sound of this idea. Here's to the Smash Lab!

 

[~ Gheb ~]

I read through the stuff Kirk wrote and it looks good. That means that in DKs case, he never get's thrown fast enough and his hurtbox is so big...