The source is the Yoga Book Hyper. I don’t personally have the book, but I know someone who does. If you know someone who can translate Japanese, that would help a great deal with clarifying all my posted info, as well as other data like Overhead hitstun/blockstun.
Edit: I updated the 1st post with my source.
Thanks! I have a copy of the YBH and the truth is I haven’t really looked at all the Japanese text (just the frame data pictures). I’ll check it out, and can even scan particular pages if that would help.
I agree that we really need someone who can translate japanese to verify stuff. There is a thread in the ST forum for YBH translation requests, but the focus tends to be on the character matchup info. And that thread probably hasn’t been posted to in quite a while too…
The hit/blockstun data is on page 173. There’s also a box saying “Hit Stop” is 14 frames…whatever the fuck hit stop is. Maybe the amount of time when you can’t get thrown after getting up? Or…?
I can mostly only read the Katakana, so I can tell which box is which data, but I can’t read all the specifics or the notes below each category. :sad:
When you land almost any attack, the two characters stop moving for a very short while for dramatic effect, then resume moving (at which point hitstun / blockstun happens). That’s hitstop.
Hitstop affects the game in two ways. First is that it gives both players a bit more time to react to what’s going on.
Second is that hitstop only happens between the two characters involved, but not any projectiles not involved in the hit, which continue moving. That has saved my Honda innumerable times, as my jumping fierce’s hitstop allows the fireball to smoothly sail two pixels under me before I land.
A variant is that when your opponent is in hitstop from your fireball, you’re still free to move anyway.
The 20 frames for non-knockdown specials is listed at the end of Yoga Book Hyper (I don’t believe it’s been scanned yet since it’s all text). The wakeup rates aren’t in there; they were calculated by T.Akiba for his site (he was also involved in YBH, although the frame data calculated in the book and on his site aren’t always identical).
Hit stop doesn’t play any part in frame data BTW. It’s just a constant extra 14 frames thrown in between your normal frames.
I was wondering why the Japanese mixed it up and use crouching MP instead of MK (specially shotos), now I know 
Right, that is my understanding too. In most cases (other than fringe cases like the one Thelo described), it nets out since it affects both players.
So, I’m looking at that section and the numbers I see amongst the incomprehensible (to me) japanese are 14, 11, 16, 20, 5, and 6.
The 5 and 6 are a little bit worrying, what do they refer to?
Anyway, I’m going to scan and post it tomorrow so we can all at least have the same point of reference.
Uh, worrying? It simply states there’s a 5 frame window to normal cancel into specials and a 6 frame window to normal cancel into super combos.
Worrying to me because I didn’t know what it referred to. But I do now - thank you sir!
Sweet, thanks. You truly are a benevolent dictator. 
jk Also, isn’t 80+ frames of blockstun for a blocked Shoryuken excessively high? It seems too high to me. I’m gonna have to start tick throwing after Ken’s blocked Jab Shoryuken. LOL But yeah, a Japanese translation would help greatly.
Also, isn’t there anything about Overheads in the YBH?
That wouldn’t be the ST Wiki thread would it? Cuz I didn’t see any requests for Japanese translation in there. Tho, a lot of the posters seem to know Japanese players that they could email. If it’s another thread, lemme know, and I’ll post my info there as well. If it IS that thread, then I’ll post there too.
Well, the hitstun/blockstun off a cr.Strong/cr.Forward are both 16 frames. The reason you see cr.Strong more often is cuz it recovers faster and therefore gives way more frame advantage than a cr.Forward. Not to mention it has a pretty decent offensive hitbox for a medium attack, the vulnerable hitbox is pretty far back and is therefore hard to trade with.
Fascinating, it appears that they set hitstop to work like a system in physics. Hitstop occurs only between two masses/forces. Which allows the opponent to move freely after a fireball hits. :tup:
More YBH stuff
So that we all have the same content to refer to, I’ve scanned pages 173-199 of the YBH and built a PDF which I’ve made available here. I hope you like Japanese :wonder:
FYI, the YBH basically has four sections:
Character-specific frame data (which I’ve already scanned and made available here).
Character-specific strategies. Lots of japanese text - see here for translations of some of that text.
A small section that contains the hitstun and blockstun numbers and some other content - this is pages 173-177.
An addendum that has information on some famous players, an interview with Daigo, and some other miscellaneous stuff.
So, I believe all the information regarding hitstun, blockstun and such like is in pages 173-177. I looked at the Ryu and Ken character-specific strategy text and they don’t look like they have any numbers in them that would be frame data.
The character-specific frame data that I scanned in and referenced above contains the startup, hitting, and recovery frames for every overhead (Ryu’s, Ken’s, etc.) I’m sure there is information in the character strategies text regarding the use of overheads, but it’s in Japanese.
My guess is that you calculate frame advantage for overheads just like any normal move. e.g. Ryu’s towards+MP overhead has 16-(2+3+10) = +1 frame advantage.
Here is the thread in the ST forum that talks about YBH translations: http://forums.shoryuken.com/showthread.php?t=124818
Yet another reason to thank NKI…
That is some pure awesomeness right there! Thanks a lot Geo. That really helps the community. Now we just gotta get somebody to translate it.
I’m loving the pics in the first post!
Here are a few comments offered in good faith…
– Where it says: “The standing/crouching blockstun for knockdown Special Attacks (eg Shoryuken) and Super Attacks vs each character are dependent on that character’s getting up speed.” It seems to me that the amount of stun generated by an attack that hits, is not blocked, and that knocks down doesn’t matter - it’s just a knock down and any stun gets washed out. For that reason, I think this reference to getting up speed should be dropped.
– Where it talks about how the stun data for non-knockdown supers is an “educated guess” - I’d need to go check, but what supers don’t knock down? And how are the numbers listed calculated - is it literally just a guess?
– I’m not sure whether knowing the amount of stun for jumping normals has much utility. For jumping attacks you’ll still have N number of frames after you hit until your character lands, where N is variable and dependent on how late you push the button (unless you’re a robot). Maybe someone can enlighten me?
You are quite welcome sir. Thank you for all your efforts!
The amount of frames it takes each character to get up from knock down is connected to how many frames they will be in blockstun after blocking a knockdown Special or Super. Example: Boxer will be in blockstun longer than Claw after blocking a Shoryuken. I think this connection is very important and should be left in there.
You’re right that all Supers knockdown, but the 1st few hits of some Supers do not knock down right away. Example: Ryu’s Super or Blanka’s Super. There are situations in the game where a character like Sagat can be hit during his Tiger Shot recovery by the 1st hit of Ryu’s Super, but will still have enough time to block the rest. Same with Blanka’s Super, Chun Li, Fei, etc. It is for these situations that I’ve specified the hitstun data.
According to the person with the YBH, who helped me translate those pages you uploaded, the data for Super Attack hitstun/blockstun is not in the YBH, but that he thinks that it’s the same as Special Attack hitstun/blockstun. Just like you, I think it’s important to get a 2nd opinion, and I’m sure he would agree too. The more people that can confirm this info, the better.
It definitely helps in those situations where people want to know whether “taking the hit” in order to “sac throw” or counterattack is a viable strategy. And according the data, it is, since hitstun off a jumping attack is much shorter than blockstun. It definitely helps my Dictator when trying to get that pesky Ken off my back with his j.Roundhouse, Knee Bash shenanigans. You’re thinking of the data in terms of the offensive player, but it’s more useful for the defensive player who has to block and then figure out his options.
I posted in the YBH translation thread, so hopefully we’ll get some use out of it. Thanks again for everything Geo. Yeah, those pics are cool, but hopefully they’ll help ppl get a visual for the data.:woot:
First hit of Blankas super knocks down in HDR.
You’re right, and so does Fei’s actually. I meant 1st few hits of Supers like Ryu or Chun Li.
The “person” that I’ve been mentioning as my YBH source is none other than Ganelon. I wanted to originally give him full credit for this thread, but I wasn’t 100% sure if posting that info would be an infraction or not. I’ve since been reassured that it’s fine. So without further ado, Ganelon deserves 100% credit for the frame data, as well as all the info in the 1st post. This thread wouldn’t exist without him. Thanks for all your help Ganelon! 
:tup:
Edit: Updated 1st post with my source.
Excellent excellent excellent thread. :rock:
By awesome coincidence, I’ve actually been pouring over the YBH and experimenting with frame stuff in training mode for the past few weeks!
I tried eyeballing command normals and it seeeeems to me like the most obvious guess is correct: it uses the 11/16/20 value appropriate for its button input.
I also eyeballed two-hit normals just to double-check the obvious guess, and it again seems to be correct: just like any other combo, stun is “reset” by each successive hit, and the only stun that matters at the end is that of the last attack. Therefore, in the case of two-hit normals, assuming both hits connect, counting up stun length (for calculating advantage/disadvantage) begins at the beginning of the active hit frames of the second hit.
I don’t have any way of being certain about this stuff, but it makes the most sense like this too.
edit: It sounds like I may be wrong about the command normals, hahah. Please see Ganelon’s post below. v v v