Official Frame/Hitbox Data Repository

[Cia]

I could do jigglypuff after I've learned a bit more about how this program works. I can start in the middle of this week. Got a lot goin on atm.

=D

 

[DtJ Jungle]

I can do peach after i get this shiz loaded and what im doing. ill probably do the bulk of the work after i move into my apt next sunday.

 

[Indigo Jeans]

Just to safeguard myself against any wrathful purge-threats:

Doing ROB, ZSS, and TLink, but I won't be able to get the data until early-mid September for ya.

Also Verilllllllllll

 

[Amazing Ampharos]

I guess we're going to be doing every character from scratch? If other people are going to actually do stuff, I'll start with G&W and then go down the line on the PT Pokemon.

I can really do anyone and am obviously extremely familiar with PSA and such. I'll probably eventually do ICs because I don't trust anyone else to handle Nana properly.

 

[rPSIvysaur]

I'd love that AA.

 

[phi1ny3]

So w/ Lucario the basic frame data is pretty straight forward, and I have all of it gathered minus smashes/chargeable specials (aka Aura Sphere) from the past, although it might need a little "flattening out" so-to-speak.

My one beef is when I have to record growth or shield advantage. Should I just take the lowest cap based on lowest aura then later on max aura by virtue of percent alone? Or should I take the extra step in accounting for the stock modifier as well? I would normally just give a basic formula to "plug in" to put in the correct data, but I worked with a couple of other frame testers in the past for advantage/shield for one move, and the result was... surprising. For example, our first move we tested, dtilt, started with a lower advantage, then became better in frame advantage with more aura. However, it actually retrogressed according to the tester for dtilt at a certain point in the aura modifier.

gah, this is both exciting and perplexing. I might need help with the controller mapping part, but I have a decent handle around regular PSA stuff.

 

[rPSIvysaur]

Don't forget to add the frame data to the Google Doc guys. >:U

To get back to you on that Phi1, I'm going to have go figure out how aura works first.

Oh and heres the characters that no one has claimed yet.
Needed:
Diddy Kong
Donkey Kong
Fox
Ike
Kirby
Bowser
Link
Luigi
Meta Knight
Ness
Olimar
Pit
Samus
Sheik
Sonic
Wario
Wolf
Yoshi

 

[KayLo!]

I'll try to start Zelda this week.

 

[Luxor]

Falcon's basic frame data is all done and formatted correctly now except his Fsmash. It's really weird- I looked at Lucas' and the different Goto's for that and they made sense to me, but I can't find where Falcon's Fsmash's Goto is sending me. It's 121A8, but I can't find that handle anywhere, like I could do with Lucas'.

 

[adumbrodeus]

Pit and MK.


I'll get to this as soon as I can, but I need my comp working again first.

 

[Isatis]

Olimar since I'm working on his hitbubbles anyway. I'm really sick so after I feel better I'll start

 

[Mr. Escalator]

This isn't something I know how to do, but I'll grab Ness in the meantime under the assumption that I'll learn it as I go.

I'll probably be asking around for help when I inevitably get stuck.

 

[leafgreen386]

My one beef is when I have to record growth or shield advantage. Should I just take the lowest cap based on lowest aura then later on max aura by virtue of percent alone? Or should I take the extra step in accounting for the stock modifier as well? I would normally just give a basic formula to "plug in" to put in the correct data, but I worked with a couple of other frame testers in the past for advantage/shield for one move, and the result was... surprising. For example, our first move we tested, dtilt, started with a lower advantage, then became better in frame advantage with more aura. However, it actually retrogressed according to the tester for dtilt at a certain point in the aura modifier.

I'm not really sure what is going on with that dtilt, but if it has a hitlag multiplier besides 1.0, that might be the cause.

There's also a chance that it's just a function of his aura, if that "certain point" happens to be 75%. The aura follows different growth patterns from 20%-75% and from 75%-170%, growing much less in the latter. I theorize that 75% is the point that his aura multiplier equals 1, although I've never tested it for myself. I'm not really sure what this would have to do with shield advantage, though.

 

[Luxor]

Thanks, I got Fsmash and Rapid Jab all cleared up, so Falcon is totally done.

Edit: King of Evil done as well.

 

[rPSIvysaur]

I have Lucas and Wario done now.

So
[collapse="Done"]Captain Falcon
Ganondorf
Lucas
Wario[/collapse]

[collapse="Up for Grabs"]Diddy Kong
Donkey Kong
Fox
Ike
Kirby
Bowser
Link
Luigi
Samus
Sheik
Sonic
Wolf
Yoshi[/collapse]

 

[Luxor]

I'll grab Link, but I'm out of town for about a week and a half. If anyone else wants to do it before me then go on ahead.

 

[Luxor]

Done with Link, but I could use a double check on his grab data.

 

[TheMike]

Can I take a look at the data already collected?

 

[Luxor]

It's on the Google Docs and you've got permission, so go ahead.

 

[KayLo!]

Luxor's making me feel lazy.

I'll try to finish Zelda by tomorrow, but I haven't looked at PSA in months, so I'll probably need some backup if when I get stuck.

 

[Veril]

This thread is cute.

Done with Bowser except for some FAF's that need to be tested in frame advance. Anyone willing to do frame advance testing to fill in gaps in my data spreadsheets should let me know. I'm not posting these on google docs cause I'm lazy but w/e I'm giving it all to you guys. Also I'm working on getting videos of some crazy stalling techniques atm, notably with Pikachu, Pit, and Peach. Alliteration! MK was actually stupid easy, and so I decided to find out if there were a more ridiculous stall than the current perfect planking. Annnnd there is! Probably won't release this until after the next big tournament I go to, unless adumbrodeus convinces me to do so sooner. I asked several people whether I should release that data and decided against it for now. Kaylo we should talk about the pikachu tricks, I want to see if its humanly feasible and need someone willing to practice it and give me feedback.

So I've got this huge set of things I need to work on smashwise. As you can all see I'm collecting various types of data from PSA and putting it in excel sheets supplemented by my work and some 3God, Kirk, Rocketpsi and Hotgarbage data. I also need to run a **** ton of tests for Cape. Got all of the formula I need except a good kbg/dmg/∂kb formula and I set what hopefully is an efficient design for finding that out. Getting Wolf frame data for TurboEther cause he asked me in person. Writing as much as possible about DK's f-tilt cause I didn't have it memorized and someone asked me at Apex what its breakdown was and I didn't know ;_;. Dabbling in ICs AT breakdowns, etc etc etc. 2 weeks... work sometimes but not too much. I wonder how much data I can get through. Probably a bunch. Gonna post Bowser's detailed frame data next cause why not? As RocketPsi has seen from the update to my spreadsheet, I'm nearly done with ALL of bowser's moveset data.

 

[rPSIvysaur]

Not going to lie Veril. You can upload excell spreadsheets into Google Docs by clicking a button. Now that's lazy.

 

[Veril]

eat a d*** lol

rPSI edit: om nom nom nom

 

[rPSIvysaur]

man, uploading them to me on AIM is harder. xD

You can even download the ones on Google Docs to excell format.

 

[KayLo!]

Kaylo we should talk about the pikachu tricks, I want to see if its humanly feasible and need someone willing to practice it and give me feedback.

Okay. I've had people over all day everyday for the past 4 days, but tomorrow I should be relatively free after 3ish.

I'll AIM you if I see you on.

 

[Raziek]

I guess I'll do Wolf's, but my job has me pretty busy atm. :S

 

[rPSIvysaur]

C'mon guys, this isn't that hard >:U

 

[KayLo!]

You have to realize that it's RIGHT at the beginning of school season...... a lot of people are really busy atm, it ain't summer anymore.

If you want it done by a particular time that badly, give a reasonable deadline.

 

[rPSIvysaur]

I know. I just want to make sure people don't forget about this.

 

[Veril]

in some cases you can literally grab data right out of my spreadsheets.

School just ended for me but I still have work and family in from California... I'll be updating that data frequently (mostly just getting all my raw non-calculated data in one place) but I'd like to see some sort of return on what essentially is an investment in this group. Also a month of being a hermit and having no life left me pretty lonely, so I'm seeing friends and **** as well pretty much every night that I'm not working (miss the pretty girls from Anatomy&PhysiologyII though ;_; ). This stuff really isn't that hard to do guys, I agree with Rocketpsi. Just, like, find the on hit effects for one move (on Lucario, Jiggs, Bowser to get a good adv window spread). Or get the tumble % for one move on those 3 characters. Or KO% with QCDI. Something.


Is "animation duration" (in the simple frame data sheets) the FAF in the absence of IASA or is it IASA-1? This is literally the only thing that is useful to me out of what I've seen posted (out of the stuff that isn't mine in google docs) so far so I'd really like to know.

 

[Amazing Ampharos]

I haven't forgotten this, just haven't had a chance. Just throwing in on that.

 

[rPSIvysaur]

Schools just starting for everyone, so I'm giving you guys a free pass for a bit until everyone is settled in.

 

[Luxor]

Is "animation duration" (in the simple frame data sheets) the FAF in the absence of IASA or is it IASA-1? This is literally the only thing that is useful to me out of what I've seen posted (out of the stuff that isn't mine in google docs) so far so I'd really like to know.

Anyone? I've been using it as the FAF. Oh, and Veril, are you ever gonna be less cryptic and mysterious about the whole Pika/Peach/Pit power planking possibilities?

Anyway, I'm doing Luigi now that I'm back home.
Edit: Luigi done, he's ridiculously fast. Obviously broken.

 

[Luxor]

Fox also done. There was some weird stuff on his Usmash, Dsmash, and Dair, so if anyone could look at those that would be great. Comments in the sheet.

 

[Veril]

Anyone? I've been using it as the FAF.

Well animation duration should be 1 under the first actionable frame IF there isn't an iasa present. The fact that people might be use FAF and duration interchangeably is one of the many reasons I don't use other people's frame data. Seriously though FAF for moves without IASA is one of the only things that we can't get extremely easily from PSA so its pretty important that animation duration is consistently FAF-1.

Oh, and Veril, are you ever gonna be less cryptic and mysterious about the whole Pika/Peach/Pit power planking possibilities?

Why?

 

[Isatis]

half-done with olimar, still have to write some history essays so i'll probably be done in a few days

 

[Luxor]

Alright then, Veril. Also, as far as conventions go, is cooldown first frame w/o hitbubble through FAF/IASA as in the OP or first non-hit frame through FAF -1? i.e., if IASA/FAF is frame 6 and the move hits on frame 1-2, is cooldown 3-6 or 3-5? I'm leaning towards the latter since the FAF isn't really "part" of the move.

Having to make all these assumptions about the implicit conventions for frame data is the hardest part for me. I'd love it if they were all codified together. I'll have to fix up all my completed sheets to match.

@Why?
It's been a nagging curiosity at the back of my mind.

 

[Mr. Escalator]

I've got time today to work on this; expect Ness' frame data to be done. Well, maybe except for animation length and things that require out-of-PSA resources. Those may very well still be done today, but no guarantee's I'll instantly understand how to gather them.

I'll edit this post when I'm done (I'm not starting right after this post, just keep an eye out for this edit).

1st edit: Finished Animation Durations for Ness. Surprisingly easy to do; moving onto the rest of the frame data.
2nd edit: Continued filling things out. I'm calling it a day, however, but I'll easily be done tomorrow (Today, technically) with the amount I have left.
3rd edit: Busy yesterday, but finally done fore the most part. All that's left is Ness' Usmash and Dsmash, which are article based and thus make no sense to me.

 

[Veril]

Duration: FAF based on the IASA or animation length - 1

FAF is a more versatile term and I like it over duration because I can use it to describe situations where there are variable amounts of endlag (so like... every aerial and almost every jab and many specials...). First actionable frame is immediately useable to calculate advantage. The "duration" term is something I will use, and possibly something you'll want to put in for public frame data, but from a technical perspective duration I suggest recording First actionable frame.

The exception to this would be in the case of pummels/grab releases where I'd use duration for the actual pummels and FAF for the release.

I don't use "cooldown" as a term at all actually. Endlag > cooldown as a term imo. Its much better to record data in spreadsheets in such a way that you can automate formula, so having 3-5 is useless. What I would do instead is calculate endlag by having two columns devoted to hitframes (Hit1 column = first hitframe, Hit2 column = last hitframe) and if I want to my data to poop out lag values I'll have 2 corresponding formula columns set to Duration - hit1 and Duration - hit2 respectively. This will give you the range of endlag.

To use your example, if a move hit frame 1-2 and had IASA frame 6, it would have 3-4 frames of endlag.
Column 1: Hit1: 1
Column 2: Hit2: 2
Column 3: FAF: 6
Column 4: Endlag(min): FAF - hit2 - 1
Column 5: Endlag (max): FAF - hit1 - 1

I don't record endlag in my data sheets because I calculate advantage based on the first actionable frame for both characters, and any lag calculation is incorporated into those formula rather than waste space in the already excessive raw data sheets I have.

I support the use of the term endlag over cooldown though, and because I'm hella stubborn I will not be using the term "cooldown" in any data I post.

 

[Amazing Ampharos]

So, I'm starting to get to work for real, and I'm noticing some issues here. Here are my suggestions:

1. Remove the "start up" field. It's totally worthless. "Hits On" completely covers that information.

2. Remove the "Cool Down" field as well. Instead, put in a "Hitbox duration" field.

3. Some moves hit multiple times, and this format is very poor for communicating their real frame data. For instance, an important point in G&W's frame data is the gap between turtle hit 4 and 5, and there's just no way to express that with this format. I suggest a convention in which multi-hit moves have their various hitbox data expressed via commas. So for a move that hits 4 times, I might express "hits on" as 7, 10, 13, 20 and hitbox duration as 2, 2, 2, 5. That would tell you when each of the four hitboxes comes out and how long each one lasts. You'd be able to glean things such as the fact that there's single frame gaps in some places and a long enough gap to maybe do something between hits 3 and 4.

4. We need fields for invincibility and super armor. These aren't that common for normals and grabs, but they do come up and will be even more significant as we move to include special moves, evasive techniques, etc.

5. The way smashes are handled is just poor. Rather than automatically grouping by sub-action, smashes should just be consolidated with an extra field for "charging frame" that tells you which frame of the smash is the looped frame if you choose to charge. All other smash frame data should assume no charging, and charge release and the like could be calculated by viewers by simple subtraction.

6. Animation duration is useless information. All you need is IASA. If a move "has no IASA" or has an interrupt point after the end of the animation (DK and Ganon fairs are the only two examples I know of), the IASA is by definition the animation duration. We can make this into a single field that contains the gameplay useful information instead of two fields of game-internal information.

Maybe other stuff, I haven't looked at everything, but since it's a blank page, I'd kinda like to remake G&W's spreadsheet completely and use it as an example for the kind of format I'd like to do. I have no problem reformatting the work of people who already did stuff, but I think we'll get something far more useful if we put in some good conventions now.

So, permission to massively rework G&W's page as a demonstration?

As an aside, is there any reason every character has to have every field? I'd like to remove pointless rows such as jab2 on G&W when he doesn't actually have one. Is there any disadvantage to different characters having different numbers and types of rows depending on what moves they actually have?