I guess using the parry by reaction is impossible. It seems it’s only anticipation and option select. How to option select with it? Just do forward/low and input a move?
How much frames it takes for a human to react for a hit confirm? Overall i mean. 16-18 frames?
Help me, thanks.
It depends on the move. You can parry on reaction certain things if they’re slow enough, any move with 3 frames or less of startup you can only parry in anticipation. As for how many frames, I would think that would depend on the person’s reactions in general. Some people can (or claim to anyway) be able to hit confirm from c. MK, but I know I certainly can’t.
Ken’s cr. mk has 4 active frames which should roughly translates into 0.06 seconds.
Average human reaction time is 0.25 seconds for visual stimulus and 0.17 for audio stimulus.
Correct me if I’m wrong but this means Ken’s low forward is impossible to hit confirm for a human being unless I’ve overseen something and it’s possible to cancel the move late into its recovery frames, which I didn’t find any info on.
Where do I find comprehensive and extensive frame data for 3S?
esn’s site: http://ensabahnur.free.fr/BastonNew/
when you choose a move, click on the cancel button for the cancel window table. sometimes the cancel window does extend past the active frames. its most noticeable for moves that “feel” easier to hitconfirm than you would expect, like alex’s close mk.
thats dps and barely anything else lol. it’s more like 16 minimum. besides projectiles, late-hitting uohs and moves that travel predictably to telegraph themselves, parries arent done on pure reaction.
Thanks a lot for linking me that site guys.
So the cancel window is 9 frames, so according to this it’s still a read and not a confirm if people claim they hit-confirm cr.mk.
I don’t know how everything interacts, but I’m pretty sure you get some additional time due to hit-stop.
There’s also this site: http://gr.qee.jp/01_3rd/index.html. Might not be for you since it’s in Japanese, though you should be able to kind of figure out what is what on there.
There’s most shoryu type moves, some normals, almost all supers, Akuma’s Tatsus, and probably some other things I’m forgetting. I just mentioned that since I didn’t know if there were certain things he was trying to parry on reaction but couldn’t since it starts too fast.
You’re reading whiffed move data. ‘Hitbox options’ -> Types -> click on H.
(D is default (whiffed move with sprites) view, others are respectively Whiffed/Hit/CrouchHit/Block (all without sprites))
Ken’s cr.mk has 13 frames cancellable, Chun has 18 (but with a hole of 3 frames in it).
don’t be surprised to find people claiming they can. i think the confusion arises from kind of stretching the meaning of the term confirm. its certainly not in the same category as true confirms like short short or crouch strong
Stressing the education are we
Sorry to sound like a grouch but i think this question about hit confirms pops up like 5 times a year!
I’ll make a frequently asked questions thread soon cause this is too frequent!
^I was sort of thinking the same thing. There needs to be a permanently stickied thread on parry mechanics or something like that
Why do people get so attached to some number they read about describing what human beings can’t do? Any clinical information such as a minimum human reaction is not a hard rule, it’s just a best guess given the group of people they tested and the test circumstances.
We obviously see this to be true with the crap people keep throwing around regarding humans not being able to recognize one frame or less of delay. If you play arcade or even OE on CRT regularly then play OE on the lauded low lag Asus monitors you immediately notice a difference.
Not confirming a cr.mk 100% or it being below some highly generalized time threshold doesn’t make it unconfirmable. I very clearly hit confirmed a cr.mk on arcade a couple times yesterday and watched a few other folks do the same. When I did it I wasn’t looking for them to be standing or any other trick. Just looked for the block or hit or parry and acted according. None of us are Deshiken or Kuroda.
You can see when people confirm a cr.mk usually. They cancel the super just a little later than normal. It’s one of the hardest confirms but let’s not limit ourselves unnecessarily.
I think when people say not a true hit confirm for cr.MK they mean that you have a lower percent chance of confirming than cr.lkx2 or cr.mp/st.mp links.
It’s easier to hit confirm standing opponents but it becomes much more difficulty to hit confirm in tense situations against crouching opponents.
I think that’s why when people say “true hit confirm” they mean higher percentage type of moves. I guess I’m arguing for semantics and what we define as “true hit confirms”
I think of true hit confirm as something you can go into training mode and hit 10/10 on random block or random hit/block/parry. if you can confirm in a real match but not reliably in training mode, then you’re probably responding to other cues like their movement, what you expect them to do, you hear their stick or button click, etc.
Except you can do exactly that with stuff like Ken’s c.mk xx Super.
It is entirely confirmable. We trained ourselves to do it.
so upload the video where you confirm it 10/10 in training mode on random block.
Too hard on console version :V
or here, let me lay out a bit of an argument:
OE and CPS3 are very close to the same speed, so if people claim they can confirm low forward jinrai on CPS3, they should be able to do it on OE. OE training mode random block is an easy way to test if you can truly confirm something.
there’s a notable difference between “confirming” in match and actually being able to confirm it. I go through stretches of time where I can confirm low forward jinrai against Alaska players with a pretty solid success rate, but when I go to OE training mode on random block, my results are no better than guessing. I can’t really see it. so this tells me my confirms on arcade are not real confirms but rather I see/hear/feel something or I know the other player pretty well.
people point out the cancel windows but I think something no one mentions is that every version of 3s has at least 4f innate input delay. you’re not just going from the moment low forward hits to you inputting the super within the cancel window. you’re going from the moment the game shows the hit animation or plays the sound cue, your brain processes it, you hit the button to confirm within that window, then there’s another 4f of delay added on basically. like if you have 13f to cancel something and you hit your button on frame 11, super is not coming out.
I’m not saying it’s impossible but I think the proof burden is on the person making the claim then can confirm it. a fair amount of people claim it’s confirmable but I’ve never seen a video where anyone goes into training mode, turns dummy to random block, then just confirms 10 in a row with no mistakes. to me that’s the definition of true confirm. if you ask me to do that for any other standard confirm I can do it. Chun low forward, back fierce, close roundhouse, strong, shoto low short x2, Makoto stand strong, all these are something you can easily go into training mode and confirm 10/10. no problem. I don’t think Ken low forward -> jinrai falls into that same category.