Same Frame Grabs and Air Releasing (debunking myths)

I was messing around in frame advance mode today and I was able to show that a couple of "facts" floating around out there simply aren't true.




Same Frame Grabs

Lower controller port does not mean you always win same frame grab confrontations. This is easily verifiable... pause the game when two characters are within grab range of each other, and hold Grab at the pause screen. Unpause the game, and let both characters do their grab. Do this a few times and you'll see that who grabs whom is not consistent. If you're skeptical about the pause method getting the inputs right, I did test it frame by frame today, and found the same thing.

I have no idea what it is that determines who gets the grab. Over the trials (a relatively small number of trials, admittedly), I didn't notice one side even having a significant advantage in the number of grabs. I don't want to say it's random, but even if it is determined by some number of factors, it's likely so complicated that it might as well be random.

I thought this was particularly important with MLG coming up, and with the controller port rules set the way they were. So, as far as we know, there's no reason to not take the highest port, given the choice. There could be other things that controller port affects (I've heard rumours like ledge snapping), but nothing confirmed at the moment.





Air Releasing from Grabs

This "myth" is slightly less known, but it's still pretty popular.

This is not completely accurate said:

When being grabbed if you break out during the pummel animation, you will always ground release.

If you break out between pummel animations,

• if you input Jump while struggling, you will air release
• if you do not input Jump while struggling, you will ground release

The duration window for this "pummel animation" was dependent on the character doing the pummeling (speed of pummeling, hitstun of the pummel). Certain characters have pummels that can always force a ground break.

Click to expand...

It seems that this theory does describe what happens somewhat accurately, but it definitely does not provide the whole picture. Whenever I grab a Wario, if I hear them rotating the control stick, I'll stop pummeling and let him release, since he'll air release even if I'm not holding him above the ground. That still holds, but the exact mechanics are more complicated than what was quoted.

Ness when grabbed at 0% without any struggle inputs will break after 90 frames. Each input reduces that by 8 frames. I grab Ness as Squirtle, and I make sure that Squirtle inputs a pummel on the 84th frame after the grab. The first frame of the pummel animation occurs on the 85th frame, and on the 85th frame, I input X as Ness (no other inputs on Ness' side prior to this). Since the 85th frame is less than 8 frames away from the 90th frame, Ness will break out on the 86th frame, which is the second frame of the pummel.

By adjusting when I input the pummel (83rd frame, 82nd frame, etc.), and always inputting X as Ness on the 85th frame, I can force Ness to break out with a jump command at any frame of the pummel. Squirtle's pummel takes 19 frames. I've made Ness break out on each of the 19 frames, and in all 19 cases, I've seen a ground release. This means that grab releasing is dependent on more than just where you are in the current pummel animation.

I was unable to determine what it was that determined it, but I decided to examine more extensively what happens when Ness breaks on frame 2 of the pummel, and I noticed an interesting pattern. At 0%, Ness will always break out in 90 frames without any input (true for any character grabbing any character, by the way). I would always input a pummel on the 84th frame, then input a jump with Ness on the 85th frame, so he would break on the 86th frame (hitting X reduces the number of frames by 8, so hitting X on the 85th frame would result in a break on the next frame). If I pummeled 5 times (so Ness breaks at the start of the 5th pummel), he always did a ground release. If I pummeled 4 times (so Ness breaks at the start of the 4th pummel), he always did an air release. In both cases, the last pummel was done on the 84th frame.

I can only conclude that grab break mechanics are much more complicated than we previously thought, and cannot truly be predicted at the moment. If anyone can propose a theory as to why this occurs, I'd love to hear it.

One other thing I noticed is that you cannot buffer pummels. This means that it's pretty much humanly impossible to always be pummeling at the fastest rate, since you have to hit a 1-frame window every time you pummel. It also means that if you want to pummel as fast as possible, you're better off learning the timing than just mashing A.

If anyone wants to follow this up and try to figure out the exact mechanics, contact me and I'd be happy to share my limited data with you.

Interesting, disagreements over ports should cease then?

Since the sample was possibly to low, others should do additional tests.

Mota said:

Interesting, disagreements over ports should cease then?

Since the sample was possibly to low, others should do additional tests.

Click to expand...

Not cease entirely, since explosives still matter with relation to grabs but more or less yeah.

How is that paragraph untrue at all?

the obvious conclusion is that squirtle cannot 100% force a ground release. Lets see some video proof of you disproving any of those three statements first because ive tested this stuff rather extensiveley myself and came to those 3 conclusions a long time ago and dont see any reason to start changing the way I see it...

Whether its more complicated or not is irrelevant, that doesnt make what we currently believe about grab breaks a myth

Yes, the second part of the OP is misleading.

Yeah Mota, additional tests would be great.

C~Dog said:

Not cease entirely, since explosives still matter with relation to grabs but more or less yeah.

Click to expand...

Yeah, if there's no Snake, Link, or Toon Link (I think those are the only explosives that are affected by port?), port shouldn't matter, unless there's something else that it affects that we don't know about.

DJBrowny said:

How is that paragraph untrue at all?

the obvious conclusion is that squirtle cannot 100% force a ground release. Lets see some video proof of you disproving any of those three statements first because ive tested this stuff rather extensiveley myself and came to those 3 conclusions a long time ago and dont see any reason to start changing the way I see it...

Whether its more complicated or not is irrelevant, that doesnt make what we currently believe about grab breaks a myth

Click to expand...

I'll try to clarify... I suppose it was a little unclear. First of all, could you show me what you believe is the correct mechanic? If you've done extensive testing I'd like to hear what your conclusions are, because I can't come up with a system that models my results perfectly =\ The quoted paragraph seems to model the mechanics somewhat accurately, but I have no idea how well. I never claimed it to be anything game-breaking, and at the moment it's not even going to change the way I play.

Squirtle's pummel consists of 19 frames. I made Ness break the grab with a jump input on every one of the 19 frames. According to the current theory, there would be some frames where Ness would air release. In the case of all 19 frames, I have managed to get a ground release. There's obviously more that affects whether he air releases than just the current pummel. It seems like it might have something to do with how you are pummeling throughout the whole grab, based on this:

T-block in the OP said:

I decided to examine what happens when Ness breaks on frame 2 of the pummel, and I noticed an odd pattern. At 0%, Ness will always break out in 90 frames without any input (true for any character grabbing any character, by the way). I would always input a pummel on the 84th frame, then input a jump with Ness on the 85th frame, so he would break on the 86th frame (hitting X reduces the number of frames by 8, so hitting X on the 85th frame would result in a break on the next frame). If I pummeled 5 times (so Ness breaks at the start of the 5th pummel), he always did a ground release. If I pummeled 4 times (so Ness breaks at the start of the 4th pummel), he always did an air release. In both cases, the last pummel was done on the 84th frame.

Click to expand...

In both cases, Ness breaks out 4 frames earlier than he would have without input. In both cases he's breaking out on the second frame of Squirtle's pummel. But in one case, he's always ground released, and in another he's always air released, and as far as I can tell, the only difference in the cases is the number of pummels beforehand. It's a very small sample size again, but the fact that both releases have happened shows that there is more to it than just the current theory, no?

Perhaps myth was the wrong word. The current model seems to be good enough for practical purposes, but that doesn't mean it's completely accurate.

So... Even more of Brawl is random?

This makes me sad. ;_;

I doubt it's random, but it's probably so complicated it might as well be random. As I said, I don't think this will change how anyone plays, since the current model seems to be working all right for practical purposes.

When you did the grab thing you used the same characters? Or did you test it with characters who's has different grab speeds?

teluoborg said:

What about ledge grabbing priority ?
It's said here that lower port guarantees auto-snapping priority.

Click to expand...

Ah I knew I read that somewhere. Thanks for digging that up.

K Prime said:

I thought the first part was already known. But yeah, it's random from what I read before.

Click to expand...

Yeah I'm definitely not the first one to discover this, but since I'm still seeing people everywhere claim it's due to port I thought I'd get it out there.

Big O said:

I'm pretty sure your problem is pressing jump for all 19 frames. Doing so counts as holding the jump button. The fastest you can input consecutive button presses is every other frame.

Click to expand...

Urgh that's me being unclear again. I'm not inputting a jump on every frame... I'm inputting one jump command to perform the break, and I'm adjusting when I pummel so that he breaks on a different frame of the pummel each time.

I'll try explaining one more time lol. Ness when grabbed at 0% without any struggle inputs will break after 90 frames. Each input reduces that by 8 frames. I grab Ness as Squirtle, and I make sure that Squirtle inputs a pummel on the 84th frame after the grab. The first frame of the pummel animation occurs on the 85th frame, and on the 85th frame, I input X as Ness (no other inputs on Ness' side prior to this). Since the 85th frame is less than 8 frames away from the 90th frame, Ness will break out on the 86th frame, which is the second frame of the pummel.

By adjusting when I input the pummel (83rd frame, 82nd frame, etc.), and always inputting X as Ness on the 85th frame, I can force Ness to break out with a jump command at any frame of the pummel. And what I'm saying is that I've done this for all 19 frames, and managed to get a ground break for every one of the 19 frames.

AlMoStLeGeNdArY said:

When you did the grab thing you used the same characters? Or did you test it with characters who's has different grab speeds?

Click to expand...

I only did extensive testing with Squirtle grabbing Ness, since I was trying to find the window in which Ness could air release. Testing multiple characters isn't necessary when one case disproves the current theory though.

good thread. clarifies a few things and confuses even more things lol.

That was my intention =x

I added some stuff into the OP... hope it clears some stuff up.

Hmmm....this is interesting. This makes Air-releasing even more important.

Port still matters for Bowser's claw, right?