Blocking xups everytime?

Classic Fighting Games Super Street Fighter II Turbo
61

I agree with you.

Well, we naturally can’t know if he uses a macro or just has the muscle memory down. Or showing that online is the one main objective of his life. But I do not think it is a game changer.

You got it right: you have to walk for 2 frames and immediately enter ↓ ( or :arrow_lower_right:, :arrow_lower_left:) + Strong. It is really hard to get that right even if it does do more damage out of all other options. I do not think it really matters if one online player from the other side of the world gets it right or not. We already know how hard it is and what other options we have, so we all can decide if we will mind trying it or keep on working on the several other aspects of Ryu’s game.

That said, the combo only works on crouching characters, that’s why you will almost never see it being used in matches, but on desperation. You are better off using the true and tested cross-up RH->cr.Strong hit-confirm 99% of the time. That’s what almost every top-notch Ryu player uses.

The only exception I can think of is that trying such things at the beginning of the round gives you lots of meter, which we know greatly changes a number of match-ups. In addition to it, it may detach Claw’s weapon, which might be handy if you are able to push him all the way to the other side of the screen and turtle there.

PS: Fierce SRK would not work in the combo you linked. You’d also need to walk for 2 frames before that cr.Fwd. But one can use cr.Strong -> cr.RH (xx Short Tatsu for meter and position), cr.Strong -> cr.Fwd xx Fierce Hadou, or even cr.Strong -> cr.Short -> whiff rush punch and throw (OS SRK if you’ve trained that). Crouching RH adds lots of stun and stun timer, so it is always a good combo ender, IMHO.

62

Oh ok. I’m sure I’ve linked a jab DP before at the end of that combo and assumed fierce would work too.

Also something cool. I always thought the WW jab DP was 1 hit, but I did it at point blank range to Blanka and it hit twice and did like 50% damage :open_mouth:

63

Jab doesn’t work, either. Unless you once walked before cr.Fwd. Both frame and hitbox data are the same until Ryu loses invulnerability.

64

ERROR, ERROR, DOES NOT COMPUTE.

Since Papasi is Asian (I think), I’m inclined to believe him more.

65

OSBR’s Brazilian english. He must be saying something else.
Even the thread that he quoted here

http://shoryuken.com/forum/index.ph…-discussion-thread.18486/page-18#post-6154575

scroll down a few posts and you’ll see pasky’s demonstration.

You guys must have never spectate Mysterious Fighter live on ggpo.
He’s the real deal.

66

I mean aerial RH into rush punch, which is how the combo starts in Zero1’s video.

Edit: just check the follow-up:

67

Paused the actual .fba file on my cpu, to see what Ryu did on the first frame of block ; just to check if I pressed a direction in that frame or whatever. Here’s the result. Ryu’s starting to block, although the input is nothing. Thing starts to get really hard to refute… But still give your thoughts, if there’s something I miss.

http://img4.hostingpics.net/pics/671300abus.png

68

Isn’t this myth from like, ten years ago or something

69

512 Moves
In classic Street Fighter 2 there’s a small chance that you’ll get a special move simply by pressing an attack button. Since the odds of this happening are 1 in 512, these are called, "512 Special Moves."
I couldn’t spot any 512 Moves in the video myself, but if you see a mysterious Sonic Boom or Sumo Missile this could explain it.

512 Guard
In the same fashion as the above glitch, there’s a small chance that you’ll automatically block even if you weren’t holding back on the joystick. This can happen even if you were recovering from a move, too.

Source:

Maybe 512 Guard is still around in ST ?

edit:

I know these two are still in ST

Okiseme Throws
This one was unfamiliar to me so I can only offer conjecture:
When a character gets to their feet after being knocked down, there’s a brief window where they are immune to throws. If you attempt to throw them at a precise moment, however, the game will sometimes (randomly?) ignore this immunity. See pt.12 @ 8:48 for a clear example.

Blanka’s Semi-Unblockable Bush Buster
Instant (aka zero frame) attacks have a 50% chance of being unblockable if they’re done close enough. It’s like the attack is so quick that your opponent doesn’t have time to put their arms up to block. Take a look at pt.10 @ 0:44 to see an example of this.
This phenomenon persisted all the way through SSF2T. Here’s a list of zero frame moves:
WW Blanka - vertical jump fierce
Any Blanka - horizontal and vertical balls
HF, SSF2, O.ST Ryu and Ken - air hurricane kick
Akuma - air hurricane kick
HF Zangief - quick lariat

edit edit:

hows that for a twist in the story lol ?

70

Yeah that 512 stuff ! Problem is, I do the exact same thing (blocking on neutral) just after that one in the picture. Odds are now 1/262144 to perform the move twice in a row. Damn, I should have play the lotery instead !

71

Wolmar, thank you for your effort. The test seems to be a good one of having the input be in a downward direction (does it not work without left/right and down?).

However the screenshot is not any better than your original video. The video would be much better if it were slowed down, looped, then paused and frame-advanced. You can do all that and more with TRUST. In particular, TRUST has the input display integrated into the emulation itself. I’m not familiar with Input Watcher or how you went about making the video, but in TRUST you can script it and share that so anyone else can get the exact same repeatable results. If you have questions, I can take you through it. I showed djfrijoles yesterday, and now he is an expert. Hopefully I can show Afro Legends too at some point.

In any case, this is NOT something “the Japanese” know about. I still doubt the theories mentioned. Please give me something I can repeat your results with, in the form of a script made with TRUST. In the past week I’ve asked TZW, Mattsun, and Sasori, and none of them knew anything about it. When I researched the wall dive, I noticed that it takes a few frames to block, ie from the frame you start holding a blocking direction until your character actually blocks (ie he is able to go into the state of blockstun).

Being able to pause, and then frame-advance, counting the frames as you go makes this possible (and convenient-- considering the amount of study we like to do for this old game, I cannot tell you the amount of joy you’ll experience when you realize that you can PAUSE the context of the game’s actual flow and animation, advance exactly and only one frame, and have time to notice BOTH the inputs for that frame AND what is happening on the screen. You advance another single frame, notice the changes, and go make a sandwich, come back all relaxed, all the tension is gone, it’s so stress-free.

Anyway, seeing that blocking may take several frames all became clear when trying to block claw’s wall dive when I tested it; when claw presses punch in the air, (to hit high, coming from his far wall, and close to the crossup line) is generally the same frame you have to start blocking-- in fact, I think you can just press block on ONLY that same frame, and let go of the stick (ie neutral)-- that means that when his claw connects, and you actually block it (you see the yellow blocking spark visual effect at the point of impact), occurs at a frame 3 or 4 frames AFTER the frame where he pressed punch in the air and you blocked (and let go of block the very next frame).

If I can see your video’s example of the opponent going from a down direction to neutral and still block a high attack in TRUST, with its integrated input display, I don’t think I could deny this anymore.

XSPR

72

Problem is, when I start trust, it reboots my cpu :confused: Plus I don’t have a partner to test that offline with…

73

You didn’t missreaded, i said it and i don’t think thats not true, i was just talking with frijoles about if that is a 1 frame thing, then it will be almost impossible to get it consistently. In fact i was writing that in this thread when neoray started to explain it so i didn’t post anything.

Even if you can get it 100% of the times, the other guy only need to change the timing just a couple of frames and the game will eat your neutral block and well you will be dead, i think thats something similar of what mao was doing against damdai, mao knew that damdai was doing the double reversal against the wall dive, so he did the wall dive a certain way so the game ate damdai inputs and got hit by it.

That’s why i agree with neoray, this is interesting but sincerely i think is a waste of time.

Sorry about the late response, i was busy last week.

74

At the very least, with the potential this holds, we should at least confirm if it’s a waste of time to practice learning it for practical competition, wouldn’t you agree? Once we know the exact mechanics or timing to perform this, and how consistently it could be used, then it will be worth giving a real judgment.

75

Of course, im not saying that it doesn’t need to be tested, experimented, etc.

Oh well i was going to wrote a long post about it but i changed my mind, in the end im not a big fan of theory fighter.

76

“Input delay” definitely makes capturing only the initial blocking frame incomplete, you need at least the 4 frames before it as well to see the input data it was actually working with.

And in case anyone misconstrues my meaning behind “input delay”, there are several (4 iirc) frames after the game registers an input and the change is visible on screen. Sometimes memory contents change before this, but generally the state is only updated the 4 frames after the input. This is true of the arcade game as well as emulation, from what I understand.

77

Doesn’t seem like it would be a viable technique to use in a real match, but certainly interesting from a technical standpoint.
But you know, I don’t even know how the fuck Boxer gets back up on his feet while doing jabs when holding down. I can perform it manually, but I have no idea how the fuck it works. (Aside from people saying something vague about kara cancelling a crouching short or something, but not really going into the details.)

EDIT:
Hm, I’ve got a theory, but no way to test if it’s true. At least I can’t think of a way.

So when you cross someone up, they can already start to turn around (that is, the sprite will go into the turning around animation) but when they get hit, they are still turned in their original direction.
Maybe because you go into the neutral animation it skips the turn around and the game gets confused somehow, as to which way the character is facing. And then somehow I guess the game decides that facing both ways means you should just block.
Not really an explanation, but maybe it’s something like this.

How the hell you can cancel the jump startup into blocking is actually more confusing to me.

78

fluxcore’s post describes what I noticed while researching the wall dive. If you see your character go from a neutral state on one frame, to a blocking state on the next frame, it’s because you blocked a few frames earlier. And you could have let go of the blocking input direction in between the time you blocked and the time the game starts displaying it. That also takes into account that there is something TO block (and your character is not otherwise just walking back/forward).

If someone believes this is possible- to block by pressing neutral, please show us by using something with integrated input display like TRUST.

Sorry to hear that. What OS are you using? Please tell me details next time on ggpo, message or post in TRUST thread.

79

Don’t forget about emulator and USB lag. It could be that you pressed a direction 2 or 3 frames ago, and it takes that time for the game to do anything. In other words, you pressed a direction and let go, 2 frames passed and then the move came out because of the delay.

Until this is proved on a CPS2 board with a programmable pad, I don’t believe it

Also to test the 512 thing, you could enable turbo on a TE stick (20 Hz right?) and hold it down for 52 seconds. That will give you over 1024 inputs which in theory should give you two random special moves, right?

80

If someone manages to get this working relatively consistently with a programmable pad on an emulated version then ill go buy some and test on my board…

Edit: well, if thy have a convincing video… … What wolmar’s showing isnt convincing enough…I’d still also have to work out how to activate the programable pads at the same time… (do they exist with 2 ports for programming both players…?)