Tick throw question

[oldschool_BR]

oldschool_BR, which programs exactly did you use? I tried replicating this with Nebula and Kawaks (which I know can use macros), but none of them seems reliable. Nebula’s frameskip skips a bunch of frames at once instead of a single one, for starters.

But even more worryingly, in both emulators, I tried using this “Perfect tick” macro with Zangief, so he would do low short -> normal fierce throw (in Zangief mirror), but the game sometimes makes that an actual tick throw, and sometimes makes him do low short -> command hop. In other words, the macro’s timing seemed variable. Yes, I set the game to Turbo 0. It seems that both of these emulators have extra variance with their macros, so you’re probably using another one I’m not aware of?

I have used Nebula’s macros. More specifically:[LIST=1]
[]used the Japanese version;
[]set the speed to fixed turbo 3;
[]created a file called ssf2xj.mac in the MACRO directory;
[]because I do not know the correct preamble so the emulator recognizes characters, I have used the same one from SSF2, but both characters’ macros default to Ryu’s;
[]set Ryu’s macros to the ones listed;
[]set both macros to the same button;
[/LIST]
I have just noticed the timing is not strict, too. For the first tests, I had used punch, release, repeat x 4 to emulate a turbo on the macros that would not activate hands, so I could get a reversal on the tick almost every time. Today, on the other hand, I have tried using a single input to see if it would be OK, and it would not, at least not for Zangief. I have spent a lot of time trying to get a descent tick into SPD macro, and it would not work. I’d get green hands, jump or hop. I could only get an SPD by cancelling a cr.jab or right after it, but then the enemy is still in recovery animation.

For an oicho tick like

Macro1Name=Oicho tick
Macro1Move=D+4,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,F,F+D,D,D+B,D+B+3
Macro2Name=tick escape
Macro2Move=D+B,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D+F,F+3

Ryu would escape almost every time.

what about a test using normal throws vs normal throws?

If I set both fierce punches to the same button in any emulator, it seems random who gets the throw. I have used this in Nebula:

Macro1Name=tick
Macro1Move=D+4,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,F+3
Macro2Name=tick escape
Macro2Move=D+B,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,F+3

But I have realized timing varies. The first input for the defender is defensive crouch, but by the time the cr.short gets active the defender should take the hit. I have noticed he would sometimes block, which means around 3-4 frames leniency.

[Thelo]

Damn, sounds like the macro route is out of question then - too much variance makes it impossible to draw a real conclusion. Probably the only real way to test it would be to strictly use single frameskip the entire time, and to make sure to enter each player’s input manually for every single frame, ugh.

[VirtuaFighterFour]

what about a test using normal throws vs normal throws?

You can manually test any throw vs any throw in FBA. Let’s say you want to test Ken’s knee bash vs Chun’s strong throw. You’d assign f+mk to right+button on your joystick for the 1P side, then assign b+mp to right+same.button for the 2P side for the same joystick. So if you pressed ‘right’ on the joystick, player 1 would walk to the right (or forward), and player 2 would walk to the left (forward as well). Then all you have to do is get both characters right next to each other, keep them walking into each other, and then hit that one button which is assigned to both of their throws. You can test command throws versus normal throws out like that as well.

I’ve tested knee bash versus Dictator’s normal throw and it was 50/50 across the board. And I tested Honda’s command throw versus Dictator’s normal throw and it was also 50/50 right across the board. It’s a simple test when two characters try to throw each other from walking (or neutral).

Doing it from a tick is far more annoying. You have to use macros which are highly unreliable for emulators. Or you have to enter the commands manually in an emulator (have fun with that). You could get a macro controller for the PS3/360, because HDR on turbo 0 doesn’t skip frames. So if anyone has a step programmable controller for the PS3 or 360 then they can test it.

Either way I find it laughable if anyone believes that there is a ‘special reversal frame’ out of stun.

[Coth_X]

I think there are special frames that decide that whatever coth_x wants to do has a 90% chance of failing.

There may be similar frames depending on who is playing. :wgrin:

[Rufus]

Either way I find it laughable if anyone believes that there is a ‘special reversal frame’ out of stun.

Normal SPD vs Reversal SPD. 10 Trials. Normal 0, Reversal 10.
Normal Typhoon vs Reversal Typhoon. 10 Trials. Normal 0, Reversal 10.
Reversal SPD vs Reversal SPD, 16 Trials, right side 5, left side 11.
Reversal Typhoon vs Reversal Typhoon, 16 trials, right side 7, left side 9.

If a tick Typhoon is done with ‘reversal’ timing, it will not grab the opponent. (20 trials)

[oldschool_BR]

For HDR and similar games you can test your frame accuracy by doing what I call a wiggle test - with ken, go foward for 2 frames, and backward for 3, then repeat 1000 times or so. If your timing is dialed in well, Ken will be stationary. If your timing is off, he’ll most likely weave and wiggle.

…this will probably bug me until I test it myself too…:wonder:

This [media=youtube]EuLj6Vblg_A"[/media].

As for frame skipping, I would not worry too much. In fact, I am interested in the effects on actual gameplay. As I play in JP 3 and US/World 2, that’s what makes most sense to me.

Another issue that should affect timing is the operating system. Windows is not a real time OS, it is not designed to guarantee some task or program section will be executed within any time interval. Due to the system’s vulnerabilities, I have to spend resources in anti-virus and firewall programs. And there are a bunch of system tasks which either I have no control or if I close, the system reboots. I am not sure my inputs get taken into account each frame, for instance. The only perfect testing platform should be the arcade.

An example: someone has said in the Wiki or somewhere else that if both players throw at the same time, one player gets the throw and the other gets nothing: the input is ignored. I have tried Gief vs Ryu and either Ryu would be held (head crush), or he would perform a throw that would be softened. If the previous information is correct and the input is disregarded, the throw is being softened due to the input being read after the other. It should be a full damage throw.

[fatboy]

There is a thred somewhere where NKI deals with this.

Throws on a reversal frame beat tick throws. Period. You don’t get the “reversal” message to pop up on screen but it takes place at the in the same frame.

If two throws are exicuted at the same time then you have a 50/50 chance youget your throw.

There is no way that a reversal throw and a tick throw can be exicutetd at the same time by the very definition of what a “reversal” throw is.

Don’t make it any more complicated than it has to be.

[Fulaani]

so what are the properties of the reversal frame after leaving hit/blockstun? its a different type of thing than wakeup reversal where you are completely invulnerable.

[Coth_X]

There is a thred somewhere where NKI deals with this.

Throws on a reversal frame beat tick throws. Period. You don’t get the “reversal” message to pop up on screen but it takes place at the in the same frame.

If two throws are exicuted at the same time then you have a 50/50 chance youget your throw.

There is no way that a reversal throw and a tick throw can be exicutetd at the same time by the very definition of what a “reversal” throw is.

Don’t make it any more complicated than it has to be.

so what are the properties of the reversal frame after leaving hit/blockstun? its a different type of thing than wakeup reversal where you are completely invulnerable.

Yeah that is the defining question, do we know 100% whether a reversal frame happens out of stun.

We aren’t concerned with wakeup here.

Since if you miss your reversal frame on wakeup, a meaty turns it into a tick situation.

[EA_Megaman]

Yeah that is the defining question, do we know 100% whether a reversal frame happens out of stun.

We aren’t concerned with wakeup here.

Yes, it does. Otherwise you wouldn’t be able to reversal SRK a tick throw. What fatboy said is 100% correct.

Since if you miss your reversal frame on wakeup, a meaty turns it into a tick situation.

If you miss your reversal on wake-up and the opponent is attempting a meaty tick into throw, you get another reversal chance after the tick and before the throw. Or they might just go into a combo expecting you to miss it.

Also when waking up, you are only invulnerable if you execute a move on the reversal frame that has invincibility properties.

[Coth_X]

Then why are there implications that a reversal throw in that situation only has a 50/50 chance of winning?

[Shari]

Then why are there implications that a reversal throw in that situation only has a 50/50 chance of winning?

Because people like making up excuses when they screw up.

[Coth_X]

Because people like making up excuses when they screw up.

So there can’t be any validity in the claim?

[Shari]

So there can’t be any validity in the claim?

Burden of proof.

[Rufus]

Double post FTL…

[Rufus]

So, I ran several tests…
EDIT: And was confused by negative edge specials…

Test 1:
Double hit jab -> simultaneous reversal SPD or Typhoon. In both cases it looks like a coin flip.

Test 2:
Single hit jab -> simultaneous reversal SPD or Typhoon & normal SPD or Typhoon. The reversal always wins.

Test 3: (Invalidated)…

Test 4:
Double hit jab -> SPD vs ‘normal grab’ on the reversal frame. Looks like a coin flip - grab out of SPD is possible even when the Reversal Attack message comes out.

So: (EDITED)
It appears there is 1 reversal frame.
Characters cannot be thrown during reversal frames except by reversal throws.

At Wheezy’s request i did Bison vs Bison with a short tick throw.
On the reversal frame, the one that got hit always gets the throw.
On the following frame it’s a toss-up.

[oldschool_BR]

I believe there is nothing special about the first frame after getting up. The only special property the character has during all the first 13 or something frames is that he can not be thrown. The WW ticks were very powerful because a character would have to enter a neutral state before being playable again. They could only block while from CE onwards all characters can use attacks and throws.

[Fulaani]

So when do these reversal frames happen Rufus, right after block/hitstun?

In test3, are you saying that you cannot do a normal throw on reversal frame, but you can get thrown. That means that perfect tickthrows are in fact unbeatable by normal throws, something we know is not the case.

In your Bison vs Bison example, if the Bison who gets hit always gets the throw on the 2nd reversal frame (there are two now?) what happens if the the Bison who initiated the tick does his throw on the first reversal frame?

lots of weirdness.

Also why are most of your tests with command throws. Normal throws are equally important.

About wakeup situations. The T.Akiba data actually mentions the wakeup frame data as an avarage XX.5 frames. Signifying you can do stuff again (jump, throw, special move, super) at XX frames (reversal message will show up for special moves/supers) but you can’t be hit until frame XX+1. (saying that meaties will whiff on your reversal frame). Hence you are in fact invulnerable on the reversal frame on wakeup.

[Rufus]

So when do these reversal frames happen Rufus, right after block/hitstun?

For the testing with T. Hawk and Zangief, it’s after getting hit by jab. For M Bison it was after getting hit by short.

In test3, are you saying that you cannot do a normal throw on reversal frame, but you can get thrown. That means that perfect tickthrows are in fact unbeatable by normal throws, something we know is not the case.

In your Bison vs Bison example, if the Bison who gets hit always gets the throw on the 2nd reversal frame (there are two now?) what happens if the the Bison who initiated the tick does his throw on the first reversal frame?

It appears that there are 2 reversal frames, and, on the first reversal frame, it’s only possible to do special moves.

About wakeup situations. The T.Akiba data actually mentions the wakeup frame data as an avarage XX.5 frames. Signifying you can do stuff again (jump, throw, special move, super) at XX frames (reversal message will show up for special moves/supers) but you can’t be hit until frame XX+1. (saying that meaties will whiff on your reversal frame). Hence you are in fact invulnerable on the reversal frame on wakeup.

Get-up timing can be a little more involved and would require its own testing.

[Rufus]

It occurs to me that I might be getting the negative edge on the specials…will need to test a little more.