noodalls investigates JG (and some GI)

http://www.mediafire.com/?5x9y66yv6ya4dhz

This is a table looking at five different moves. The numbers refer to the timing of Taki's A hitting (i10 frame move). I used Asta's A for his standing K, and subtracted 10 off the number, just for comparison's sake.

The summary is, JG adds 5 frames of recovery to the attacker. JG causes the defender to recover for 19F. How much better off you will be after JG ranges from 5F (where defenders recovery was 19F in the first place), to

JG advantage = attacker's normal recovery + 5F - 19F

Now, the way to calculate a JG frame advantage list are two methods. One just do it the classical way,

two work out the total frames for all the different moves in a movelist. This is easily done with recording, turbo and visual interpretation. This may have additional benefits for combo video makers in terms of collision hits leading to combos.
 
Good stuff. I'm still not convinced safe JG doesn't exist though since in your screenshots the second guard is pressed on the same frame as impact.
 
The second guard marker appears on the same frame as impact. SCV displays them a frame late from what I can establish*. Guard was pushed the frame before impact.Besides, if it wasn't pretty sure they wouldn't be guarding.


To compare two similar moves, in Tekken almost all jabs are i10. Natsu's A is i10.

01234567890
A B
A B

The top row is how Tekken displays input and hit markers (where A is the input appearing on F0, animation begins on F1, hit marker appears on F10.)
The bottom row is how SCV displays inputs and hit markers (where A is the input appearing on F1, animation begins on F1, hit marker appears on F10).
 
I would still test to have the attack hit somewhere in the middle or even at the beginning of a successful JG. Right now it just looks inconclusive with room for doubt. Especially when with your successful example you can see the JG animation start up from the frame of impact. If a second G were to cancel a JG I would expect to see a JG animation that gets canceled. I actually have the equipment to test myself but I don't relish the idea hooking up my arduino to some PCBs right now.
 
There is no JG animation unless a move connects. In the single press example, is G is pressed and released ~5 frames before the animation connects, but the JG animation only appears when the attack connects on F10.
Especially when with your successful example you can see the JG animation start up from the frame of impact.
Click to expand...

That's what it does.
 
Which means your example is faulty. What you need to show is the move connect on the first frame the JG becomes active and then have G depressed at the end of the JG window. If a JG occurs and doesn't have its animation canceled it means safe JG exists.
 
Using the above example. Natsu A will hit on F10, where it is pressed on F0. The latest possible opportunity to input G would be F8 (released on F9) for JG to occur. The latest possible opportunity to guard is F9. You can't both release and press G on F9.

Within the framework of pushing A on F0, tell me exactly when I should be pressing G to fulfil your conditions, I'll record it and post it for you. Then we can settle this.

Which means your example is faulty. What you need to show is the move connect on the first frame the JG becomes active and then have G depressed at the end of the JG window. If a JG occurs and doesn't have its animation canceled it means safe JG exists.
Click to expand...
Unless I'm reading this wrong, you want me to press G on F8, release on F9. Then press G again on about F11, a time after Natsu's standing A has already hit, and see if it cancels the JG animation. I can already predict the result, but if it will cause you to concede I can do that. Otherwise, work something out as per the text above the quote.
 
That would work. But my example of a safe jg would be pressing g again just after the jg window ends. So more like f15.
 
http://www.mediafire.com/?aupzrpjbjc39lrg

Still don't understand how pressing guard 5 frames after the attack has hit has anything to do with anything, but here you go. JG animation is not cancelled out by pressing G. Two variations, one pressed on F5 like the prior examples, and one pressed on F8, the last chance to still get a JG.
 
What you've posted pretty much proves the existence of safe jg and mirrors my expected results. So yeah. Now to think of ways to deal with it.

Anyway, I hope I don't come across as being an asshole. I think you may be misunderstanding why people are tapping and then holding guard. The idea is to be able to attempt a just guard but keep it safe if you happen to miss the timing. It's not to increase the chance of one.
 
If i understand the best we can do is list all the blockstun create by moves or/and list every recovery of moves

I started with Mitsurugi but i have some trouble. How can i know exactly when a character has fully recovered of a blockstun state and how i can notice the character has finish his move ?

Example if i do a low on Natsu CPU, she block low but after she "waste time" to stand to perform an AA.
Perhaps using CE is a better choice

What about active frames ?
 
ShenYu said:
What you've posted pretty much proves the existence of safe jg and mirrors my expected results. So yeah. Now to think of ways to deal with it.

Anyway, I hope I don't come across as being an asshole. I think you may be misunderstanding why people are tapping and then holding guard. The idea is to be able to attempt a just guard but keep it safe if you happen to miss the timing. It's not to increase the chance of one.
Click to expand...


I think then that safeJG is a misnomer. It should be something like safe-failedJG. As I said in another thread, you're basically cheating yourself out of JGs.


Correct timing with no extra tap --> JG
Wrong timing with no extra tap* --> get hit
Correct timing with extra tap** --> guarded, no JG
Wrong timing with extra tap* --> guarded, no JG.

So you are reducing the risk of getting hit with the wrong timing, but you're also taking away all your successful timings. Alternatively, you could have just held guard.

* this is the point that you seemed to want to make. If the extra tap occurs too late, it won't stop your JG animation. Conversely, if you're using that timing, and failed to JG, it's not going to block either.

** this is the point that I wanted to make, that you can ruin a valid JG by adding an extra press. You will guard, but you will also lose your JG potential.
 
noodalls you assume the extra tap always occur before you get hit but i think what ShenYu is saying is they do the extra tap at the end of the 6f window. Like that they stay safe and they try a just guard anyway

See that like the absolute guard in street fighter where you can let your guard off during a blockstun but in this case they let the guard off during just guard and hold again at the end
 
c_nul said:
If i understand the best we can do is list all the blockstun create by moves or/and list every recovery of moves
Click to expand...

No. My understanding is that the blockstun created by moves is irrelevant, because it's going to be replaced by the JG recovery, which is a constant. What we need to know is how long the attacker recovers for.

The way to determine this is

- set CPU to JG
- perform the move you want to check, followed by a move of known initiation.
- record using Hauppauge
- review frame by frame using avidemux/sony vegas extc.
- If we make the point of contact frame 0, the opponent will recover 19F later
- The recovery for our move is the time of the second move connecting minus its initiation frames.

I'll post an example to show what I mean.
 
noodalls said:
Correct timing with extra tap** --> guarded, no JG
Click to expand...

You just posted proof that this works. I'm confused how you're still saying this.

Of course if you press G again before impact that would cancel your JG attempt. You're potentially reducing your JG window, however those frames where the JG window is open still means the JG is there. The trade off between having a maximised window for JG and having a normal guard with a JG attempt in there is 100% worth it. And frankly it is difficult to humanly press, release and press again in under 5 frames. So chances are you're not losing much of your JG window even if you were.
 
What I would like to know, is what is the earliest frame you can attempt JG, if you are holding guard beforehand?

noodalls said:
Interestingly, it doesn't matter whether you held or released your initial G, you can release it on F28 and press G again from F29-32 and JG will still occur.
Click to expand...
Is this it? I think it is...

If this is so, when it comes to Natsu BB, why is it so hard to regular block B, and then Just Guard the second B?

Also:

How do the frames work when it comes to Just Guarding entire strings (say, JGx2 Natsu BB)? Does JG not have cooldown when it is successful?

---

Hold G for 5 frames, release. When is the earliest time you can attempt Just Guard from this point?
 
DrakeAldan said:
What I would like to know, is what is the earliest frame you can attempt JG, if you are holding guard beforehand?
Click to expand...

I believe it's 29F from when you started holding G.



If this is so, when it comes to Natsu BB, why is it so hard to regular block B, and then Just Guard the second B?
Click to expand...

Dunno. Looks to have the same 6frame JG window.

Also:

How do the frames work when it comes to Just Guarding entire strings (say, JGx2 Natsu BB)? Does JG not have cooldown when it is successful?
Click to expand...

This is going to be very hard to determine. My guess would be there is no frame where you can't JG post successful JG, however what we need is two moves that hit 15F apart (i.e. 20F given the +5F delay of JGing). Natsu's AA is 26F apart for reference.

Conversely, it appears that if moves hit during the 19F of JG animation, then they will automatically be JGd. Cerv ws+A+B fulfils this criteria.

---
Hold G for 5 frames, release. When is the earliest time you can attempt Just Guard from this point?
Click to expand...
[/quote]
29F from when you first held G, or 24F post release.
 
Thanks for the reply.

noodalls said:
29F from when you first held G, or 24F post release.
Click to expand...
Still that much?


Skip to 2:25 (Battle 4). I get smacked, get up, have my guard up, then Just Guard. Could 24 frames have passed during that time?

Please excuse me, I have problems ballparking frames in terms of feeling.
 
You must log in or register to reply here.
Back