The Onion - Gathering and Storing our Frame Data!

[DtJ Hilt]

Important Posts:

Down Throw Hitstun on Meta Knight

Down Throw Hitstun on Snake

Frames it takes to land from with a fast fall from a respawn on most legal stages

Pummel Downthrow Discovery

Stale Up Smash Hitstun Discovery

Stale Up Smash Hitstun Dicovery Part 2

Luxor's Shield Stun Data


-----------------------------------------------------------------------

First of all, I want to thank Fino a lot for what he did for me. A few days ago my external hard drive broke. And for those of you that don't know, since my wii won't read discs anymore, I have to load games through an external source, such as a flash drive or an external hard drive. And since mine doesn't work anymore, I was out of luck. Three days later, I got a new Flash Drive in the mail that Fino bought for me.

Much love, man.

Now to kick off things for the Olimar Boards a bit, since we've died down lately, I decided to make this thread. The main reasons I've made this thread are

1) I'm able to have a log of all of the data I collect, and a place to put it, so that others are able to keep up with it, and that I'm able to have it all in one easy accessible place.
2) People will be able to volunteer to help if they would like, although this is not me necessarily me asking for help. You're welcome to volunteer yourself, but I should be fine even if it's just me.
3) Others are able to suggest small projects for me to do, or ask questions about data that we don't have that I can look into.

However, I have a few things that I want to state.

The first being, I am not going to be slaving away getting every piece of frame data that we don't necessarily need. I'm going to be focusing on what's important, while still doing small projects that don't take long. My main reason for ceasing my obsession with wanting to know and list everything is that I'm wasting time that I should be using to get better. If I'm wasting time figuring out things that are not important, that's time that I could be spending either A) improving my skills at the game in general, or B) researching something that is important. The other reason is basically.. it's too much for me to do by myself with limited help. Vocal has expressed wanting to help me with it all, and others have shown interest as well. But it's just too much for a couple of people to do. So if I'm going to be looking into things, it's going to be what's important.

That said, I am going to take requests, but don't be surprised if I refuse your request. For certain people I'll be more lenient to take requests from (like if you bought me a new flash drive lol <3) but if I feel it's not something that is important, I'm not going to deal with it, unless I just have the free time or if it's something that wouldn't take much time.

--------------------

Projects I'm currently interested in:

Start and End Data for Every Hitbox:
When I originally did our frame data, I didn't get everything. The main thing that I didn't finish, was the end data for most hitboxes, frames that sweet/sour spots occur and end in, etc. I got all of the start data for the first hitboxes of every attack, but I didn't get all of the end data. This is something that I can split up through time, and I can do it while I do other projects without it taking much time. And I feel this is important. I'll probably compile a list of what hitbox data I have, and start filling it when I have the free time.

Shield Stun Data:
This might be the first thing I start on. Shield Stun is the amount of time that an opponent is stuck in a shield after a move connects, and I feel this is something extremely important that we need to know. Especially for moves like Up Smash, that we're often able to throw on a shield and be safe against many opponents. Knowing how much frame advantage we get would be really helpful, as well as knowing how safe it is against certain characters.

Hitstun for Combos:
I'm not going to be getting hitstun for every attack. Hitstun is dependent on percentages, and differs from each attack and pikmin color. Different attacks and pikmin colors accelerate hitstun at different percentages, and this is too much for me to deal with, trying to get it for every attack. Coupled with the fact that I would have to do this in VS Mode, making it tedious, I'm going to focus on getting hitstun for moves that matter. I'm going to start with the pieces of our Down Throw Combos.

These are going to be the main three Major Projects that I'm going to be focusing on, for awhile. If anyone has anything specific that they would like to request me look into, that's fine. That's why I'm making this thread.

I'm not asking for help, but if you would like to help me with this, you are more than welcome to. I would really appreciate it actually, haha. Just let me know if you would like to help me and I can help you set up the hacks on your wii and give you a rundown on the process for whatever you're helping me with. Keep in mind that doing this can get really, really boring, which is why I usually don't expect people to help XD So my feelings aren't going to be hurt if you don't.

I'll be starting on this soon

 

[DtJ Hilt]

The first thing I'm going to be focusing on is Shield Stun data. However, Shield Stun decreases with the staleness of the move. So, only testing how much stun there is when it's fresh isn't enough, since our Up Smash is rarely completely fresh. But I'd like to get a general range of how fresh our Up Smash usually is, but this is really complicated XD So I'll try to get data for it with it in our stale list at different times and

 

[Excellence]

My question to you is: Does hit stun decrease at a constant rate per hit? For example, if I use USmash twice and it hits shield both times, if I repeat this process (not consecutively) will the hitstun be the same as it was before? Suppose both hit, do they have the same knockback? Is there a minimum hitstun?

 

[DtJ Hilt]

Well first of all, you're talking about two different things.

Shield Stun is only determined by how stale the attack is. It doesn't actually make contact with an opponent, so the move doesn't get stale from hitting the shield.

Hitstun, however, is completely different. Hitstun changes depending on the percentage of the opponent, however this is not a steady drop or rise. In fact it does both!

For an example, let's look at White Up Smash. Hitting the opponent with a fresh White Up Smash at 0% will cause 28 frames of hitstun. Up through eight percent, this amount of hitstun slowly increases, dealing 30 frames of hitstun between 6 and 8%. However, at 9%, the amount of hitstun decreases dramatically to 14%! From 9% to 999%, it stays at this amount of hitstun. This was tested on either a Meta Knight or an Olimar, I don't remember, it was a long while ago. I recently came to the realization that hitstun is weight dependent, and heavier characters do recieve more hitstun than lighter ones (based on weight, not fall speed). So it's going to take a bit more time to figure this out than I thought, haha.

If I didn't answer your question, tell me what you mean

 

[Noa.]

This is a good thread for us(you) to keep our frame data updated. Will you be placing the existing frame data into the OP of this thread?

 

[DtJ Hilt]

Probably, but I'll be reorganizing what I have, in a way, first.

 

[Excellence]

Well first of all, you're talking about two different things.

Shield Stun is only determined by how stale the attack is. It doesn't actually make contact with an opponent, so the move doesn't get stale from hitting the shield.

Hitstun, however, is completely different. Hitstun changes depending on the percentage of the opponent, however this is not a steady drop or rise. In fact it does both!

For an example, let's look at White Up Smash. Hitting the opponent with a fresh White Up Smash at 0% will cause 28 frames of hitstun. Up through eight percent, this amount of hitstun slowly increases, dealing 30 frames of hitstun between 6 and 8%. However, at 9%, the amount of hitstun decreases dramatically to 14%! From 9% to 999%, it stays at this amount of hitstun. This was tested on either a Meta Knight or an Olimar, I don't remember, it was a long while ago. I recently came to the realization that hitstun is weight dependent, and heavier characters do recieve more hitstun than lighter ones (based on weight, not fall speed). So it's going to take a bit more time to figure this out than I thought, haha.

If I didn't answer your question, tell me what you mean

You did a great job of answering my question. Is there someway that we can break this down into a forumla so that it can become a plug-in type method? Otherwise, I see this as taking a really long time.

 

[DtJ Hilt]

I'm going to try to talk to the Researchers in the Smash Lab (since I finally got in XD) and figure out and understand the formula, but yeah.. until then it'll definitely take a good while, lol.

 

[The Fluff]

3) Others are able to suggest small projects for me to do, or ask questions about data that we don't have that I can look into.

I'd like to suggest that we find out the exact percentage to pummel ratio for Oli's grab, with all the pikmin. I'd be happy to help if you were interested in doing that (I'll still do it anyways even if no other Oli wants to, lol).

 

[DtJ Hilt]

That's a really good idea, actually. The only problem with it is that I don't currently know the fastest the opponent is able to mash out, and that would be important. However, I know a thread was made on it in the past, so if I could figure that out, I think we could compile that list, I really like that idea, actually!

Another thing I'm really interested in looking into is comparing Shield Canceled Dashes vs Pluck Canceled Dashes vs Stopping an Initial Dash Normally. I originally compared Pluck Canceled Dashes vs Stopping a Regular Dash, but never an Initial Dash. And I'm not sure how long it takes for Olimar's Dash to take off, which was something Fino asked me a few days ago, that I should look into.

 

[Luxor]

IDK what this post is, but I'll post this post I wrote.

Start and End Data for Every Hitbox: idk how you'd even start, since Oli's weird in PSA. Start is easy enough in frame advance, but the end is hard... you could try timing spotdodges so they get hit at the very end, I guess.

Shield Stun Data:
http://www.smashboards.com/showthread.php?t=285502 if you didn't get the formula yet, but I think you have it already.
[collapse=Block Advantage Formula]normal escape frame on block=hitframe + SHL + sstun
SHL=Floor[(damage/2.6)*E+5]
Sstun=Floor[(damage)/3]
hitframes you have
Your escape frame: normal FAF + SHL
Total Block Advantage= hitframe + Floor[(damage)/3] - non-hit FAF

So all you need to know to get block advantage is:
1. What frame you hit on, doesn't have to be the first possible hit-on frame.
2. damage attack does, you can factor in staleness if you want
3. the FAF when you don't hit

3 is easy w/ PSA, 1 is easy in frame advance but hard in PSA, and 2 (the exact number) is hard, but a rough idea is easy.[/collapse]

Hitstun for Combos: I'd put this as low priority, since it's so variable and needs another formula for this.

Down Throw Combos: This should be a piece of cake, just test 10% intervals on Lucario/Bowser/Jiggs controlling for pikmin and SD'ing every throw. A bit time consuming, though.

 

[DtJ Hilt]

End data for the hitboxes isn't hard at all, I've actually already gotten quite a bit of it, haha. Thanks for the rest of the input, though

 

[The Fluff]

That's a really good idea, actually. The only problem with it is that I don't currently know the fastest the opponent is able to mash out, and that would be important. However, I know a thread was made on it in the past, so if I could figure that out, I think we could compile that list, I really like that idea, actually!

I'll do some looking for the thread. In the meantime, I'll start doing some testing using my turbo controller in the P2 slot, it should be pretty accurate I think. Frame data shouldn't be too important for testing, I would think.

I was thinking, do you think Oli could do a buffered Hyphen Yellow U-smash after a grounded grab release? Or rather, what type of grab release options do we have? At certain percents, pummel > Grab release follow up might be really good. I.E. 110%, character isn't in U-throw kill range, pummel three times, get them to 116% or so, Hyphen U-Smash them for kill. This should probably be in the other thread though :\

 

[DtJ Hilt]

You're going to need to mash with the control stick as well. And the C-Stick. Not just buttons.

 

[Luxor]

IIRC if you do it in frame advance, you're limited to one escape input per frame. In realistic terms that's stupidly fast and will get your grab broken out of really quick.

Also, I doubt Oli has a guaranteed Usmash out of grab release. I don't think advantage on release is percent dependent (only character dependent), and almost all characters are the same. I'd look into the Earthbound boys, though, since they fail like that.

 

[The Fluff]

You're going to need to mash with the control stick as well. And the C-Stick. Not just buttons.

Well, for whatever reason, the turbo controller I have won't work for brawl :\. It works for other GC games, but it just won't function on brawl. I'll see if I can get my sister to help me out tomorrow.

IIRC if you do it in frame advance, you're limited to one escape input per frame. In realistic terms that's stupidly fast and will get your grab broken out of really quick.

Also, I doubt Oli has a guaranteed Usmash out of grab release. I don't think advantage on release is percent dependent (only character dependent), and almost all characters are the same.

I never implied that it was percent dependent. All grab releases have the same amount of lag and are released at the same spot, regardless of percentage. I was, however, saying that it might be possible to get a kill by Grab Release>USmash when an up throw won't kill. I'm thinking grab with white pikmin, pummel (lot's of damage, lol), and then USmash if applicaple after the release.

For example, taking data from this thread, Mario is killed at 114% with a purple UThrow. However, if Mario is grabbed at 113%, he won't be at ko percentage for a purple UThrow. If he was d-throwed, he would take 8 damage, putting him at 121%.

Now, say that Mario at 113% is grabbed by a white pikmin. White pikmin do 3-4% pummel damage. That means, only 3 pummels have to be done to equal the damage from a Purple d-throw. Now, if Mario got grab released, and could then be followed up by a Yellow DACUS (since it covers ground faster, it would make it easier to land, or rather, might be the only way to land it), a Yellow U-Smash kills him between 106%-124% (Close>Far). Only about 4 pummels would be needed to put Mario into killing range, which could be done.

Another thought is that Olimar can pummel, and then use a buffered pikmin toss as a grab release option. If the opponent attempts to retreat, a white pikmin can be latched, if they jump, yellows will always get them, and depending on the character, any other pikmin aside from purple could latch onto them when they do either option. If the opponent spotdodges, I think you can get a regrab.

Oh boy, I need to do some testing :D.

 

[Noa.]

Fluff you're pretty knowledgeable.

Do you go to tournies or not? I know you probably don't judging by your post count but you should.

 

[-Vocal-]

I love you for making this thread Hilt :D I do have ONE request if you have the time to do it - do you think you could get data on how long hitstun for Dtilt lasts? I'd like to know just so I have a better idea of who's safe to Dtilt without having to fear a quick aerial as a counter before the move ends

 

[Luxor]

I know what you mean, Fluff, but the issue is that I seriously doubt you have the frame advantage to connect a Usmash before they sheild/spotdodge/roll. The only characters in this game you might be able to do it on are MK/Wario out of an air release or Lucas/Ness/Squirtle if squirtle has a bad ground release, I don't remember if its air or ground release is bad. Unfortunately, Oli is the only character in the game who can never ever for never ever air release, so goodbye guaranteed kill on Wario/MK. The only char IIRC who really gets good ground release options is Bowser since he has an abnormally large frame advantage. With Oli, your opponent can move and react to Usmash, it's unlikely to be guaranteed, and you get punished because Oli doesn't have a Bowser-like ground release.

Does that mean GR--> Usmash is bad? Not necessarily, it could e a good mixup. If you try to force it, though, the frame data will come back to bite you in the butt.

 

[The Fluff]

Fluff you're pretty knowledgeable.

Do you go to tournies or not? I know you probably don't judging by your post count but you should.

Nope, I can't get to any tournaments. Partly because I live in Illinois (closest tournament where I could get some dedicated smashers is "Don't Blink!", but that's 1.5 hours away from where I live) and partly because I don't have transportation. If I could go to tournies I definitely would. Every time I go down to Nashville to see my dad, I always check if there's any tournies while I'll be down there, lol.

I know what you mean, Fluff, but the issue is that I seriously doubt you have the frame advantage to connect a Usmash before they sheild/spotdodge/roll. The only characters in this game you might be able to do it on are MK/Wario out of an air release or Lucas/Ness/Squirtle if squirtle has a bad ground release, I don't remember if its air or ground release is bad. Unfortunately, Oli is the only character in the game who can never ever for never ever air release, so goodbye guaranteed kill on Wario/MK. The only char IIRC who really gets good ground release options is Bowser since he has an abnormally large frame advantage. With Oli, your opponent can move and react to Usmash, it's unlikely to be guaranteed, and you get punished because Oli doesn't have a Bowser-like ground release.

Does that mean GR--> Usmash is bad? Not necessarily, it could e a good mixup. If you try to force it, though, the frame data will come back to bite you in the butt.

I'm pretty sure Squirtle has a bad air release, it's what allows Marth to CG>Spike him at 0%.

I'm just saying, it could catch an opponent off guard. USmash hits on frame 8. Unless the opponent spotdodges or shields, I'm pretty sure a yellow DACUS will hit. Definitely will hit on Ness/Lucas, since Usmash's first hitbox will come out on frame 38 if buffered, and Ness/Lucas get out of a grab release on frame 40. I think it might even be possible to poke some character's full shields depending on what frame we start the DACUS. Oh yeah, USmash's hitbox ends on frame 64 after the grab release, so that'll hit spotdodgers, no matter what frame they input the spotdodge in for (assuming spotdodges are 22 frames, I think Link/Falco are 24 frames, correct me if I'm wrong).

You're right that it'll probably be just a mixup, but it might be worth going for if it potentially racks up more damage, or even kills, when a throw won't kill.

 

[DtJ Hilt]

Fluff, you're a pretty cool guy. Stick around here more.

I'll look into it, Vocal. It may be difficult because of the antenna factor.

 

[The Fluff]

Fluff, you're a pretty cool guy. Stick around here more.

I'll try .

Sister is out shopping with other sister and her friend, when she get's back, I'll be doing testing (hopefully).

 

[Greward]

thats a good idea. we definitely need to work on our oos

 

[The Fluff]

Hilt, could I create a new thread for the research I'm doing on grab releases/pummels?

 

[DtJ Hilt]

You could, or you could post what you want to in this thread, I don't mind either way. Feel free to make your own thread if you'd like

 

[The Fluff]

You could, or you could post what you want to in this thread, I don't mind either way. Feel free to make your own thread if you'd like

Alright. I'm thinking it'd be best to make a whole thread, so that way we can sticky it in the forum directory, if/when it's completed. Also, Olimar has two different ground grab releases and an air grab release, so that'll take up quite a bit of space to write everything out I think. I'll do some typing tomorrow when I wake up .

 

[DtJ Hilt]

Feel free to work on the thread, but the thread most likely won't be stickied, just so ya know.. X:

 

[The Fluff]

Feel free to work on the thread, but the thread most likely won't be stickied, just so ya know.. X:

Wow, I fail. I said stickied in the forum directory. I meant to just have it listed as one of the reference threads (like pikmin pluck percentage thread) if it ever got completed. I need sleep :X

 

[Dnyce]

I HAZ REQUEST~!

http://www.smashboards.com/showthread.php?t=205614 - mk
http://www.smashboards.com/showthread.php?t=258332 - snake

Here are examples on everything that you can get frame data on (might be one of the reasons they're on top ;o). Idk, I was just hoping you could use that as an outline - or a checklist - of things to test frame wise. I'm not saying get this all done now, but these two frame data directories literally break down EVERYTHING imaginable, not just attacks, but movements and specific options as well.

I am really interested in seeing this unfold. Glad I could help out, cause I honestly would be awful at testing/gathering stuff like this

 

[Luxor]

@Fino: The majority of those movement options are easy in PSA; if you're physically testing there's a problem. You only need to do that for a few scattered things here and there. Getting most of that movement data would take like... 15 minutes or so.

 

[Dnyce]

@Fino: The majority of those movement options are easy in PSA; if you're physically testing there's a problem. You only need to do that for a few scattered things here and there. Getting most of that movement data would take like... 15 minutes or so.

I don't know what PSA is, means, or stands for...

15 minutes sounds great though. How do I do it? If it were so easy though, I'm sure we'd have it by now.

 

[DtJ Hilt]

Just so you know Fino, those things are going to be at the back of the line, behind specific things. If you have anything specific you would like me to check, let me know.

Currently working on figuring out how many frames of invincibility we have after hitting the ground on every stage from a fast fall, at Weruop's request.

 

[Elyssa Xey Hexen]

Hilt, could I just steal you away from the olimars and help us Falcos?

 

[Dnyce]

I guess something I would like specifically looked at is the usmash>dsmash on shield. Particularly if a fresh white-usmash > dashing (or not) usmash > dsmash on shield is possible. Only reason I ask is because you had mentioned absurd hitstun on white usmash earlier. Or also fresh white-usmash on shield to grab

 

[Weruop]

im really interested in the white US stuff

 

[DtJ Hilt]

After respawning, every character has exactly 120 frames (two seconds) of invincibility. So I compiled a list of how long it takes Olimar to land on the stage from a fast fall, after respawning. Subtract the number from 120 and you get how many frames left of free invincibility we have to do whatever you want. Also keep in mind that pluck a pikmin takes 9 frames, and plucking a pikmin on the third phase of castle siege takes 13.

Final Destination   Main Stage     34

Battlefield         Main Stage     45
                    Platform       19

Smashville          Main Stage     45
                    Platform       28

Yoshi's Island      Main Stage     39
                    Platform       22

Lylat Cruise        Main Stage     38
                    Platform       24

Halberd             Moving Stage   38
                    Platform       23
                    Ship           32
                    Platform       24

Pictochat           Main Stage     55

Brinstar            Main Stage     33
                    Platform       7

Pokemon Stadium 1   Main Stage     38

Castle Siege        First Phase    36
                    Right Hill     33
                    Second Phase   55
                    Top Platform   26
                    Mid Platform   41
                    Third Phase    33
                    Left Hill      31

Frigate             Phase 1        40
                    Phase 2        40
                    Platform       24


Delfino             Main Stage     38
                    Dock           32
                    East Beach     21
                    West Beach     27
                    Umbrella Phase 42
                    Umbrella       31
                    Umbrella Water 81
                    Pillar         24
                    Roof           26
                    Street         37
                    Awning Phase   33
                    Shine Gate     20

Regarding delfino, these are all names I've given the phases. Some of them are self explanitory. Umbrella Phase is the phase with the two umbrellas and the water at the left side. Awning Phase is the other phase with shallow water and the red awning on the left side. East Beach is the beach near the Pillar Section, and West Beach is near the Awning Phase. Dock is the plain flat stage with the water on the side, and street is the section with the two walkoffs and the dip in the middle.

I didn't include the time that it takes to get to the top platforms of Delfino of the main phase, because I didn't know what to name them. Just know that the phase with two platforms with one on top of the other has 10 for the top platform and 24 for the middle one.

Lylat Cruise and Castle Siege Phase 3 are assuming that the stage is flat at the point of landing. Same is said for Yoshi's Island platform.

Also, this includes the frames of lag upon landing.

Using this, you can figure out how much time you have, and see what options you have and how much time left of invincibility you'll have to play around with afterwards. I'm definitely going to do a bit of math for a few of the stages because I think this is really, really important. Thank you for suggesting this, Weruop. XD​

 

[Weruop]

np. thanks a lot for all ur work

edit: (45 frame drop ---> pluck pluck sideb pluck pluck pluck = 115frames/120)

 

[Luxor]

never heard of PSA!? and Usmash followup question

PSA is... Well, the acronym is unimportant. It's basically a program that says "on move x on frame x x hitbox comes out doing x damage and x trajectory for x frames until you can move again frame x." Data galore.

As far as Usmash followups, I'm interested in white Usmash after your post. Usmash is so quick and whites do so much damage (and therefore shieldstun) that it might turn out to be a pretty sweet frametrap. IDK the damage white Usmash does, the frame it hits, or the IASA frame, but if someone told me those three I could spit out the frame advantage Usmash has on block.
I should post in the Oli boards moar :/ but w/e.

 

[The Fluff]

Hilt, that list is awesome :D. Now I just need to commit it to memory .

For all characters, what frame does a shield come out on? Or, does it vary for each character?

Also, Olimar's grab is *one* frame to slow for an infinite grab release chaingrab on Ness/Lucas :\. Though, it could still be possible I guess, since their spotdodges come out on frame 2, and Olimar's grab could grab them on frame 1 (both those numbers are the frames after Ness/Lucas is released, Olimar's grab is 11 frames to start up, but there's a 10 frame advantage for us vs Ness/Lucas), unless N/L can jump on frame 1 and get above the grab hitbox, or they have another move they can do on frame 1 after their release (which I didn't see when I skimmed their frame data thread, fastest move either can do is Lucas's jab on frame 1, but it doesn't hit until frame 2.

 

[Greward]

u should put some random onion pic in the op