Frame Data Project - Second Try - The Beginning

[Yikarur]

I'm bothering for this subject for a really long time, because I was really unsatisfied with the last try.
that why I will restart the project all over now.

when collecting frame data we have to question few questions.

What form?
ex. Tabular (Excel), Thread (like most boards already have), extern webpage with all data compiled, txt. datas for download, etc. etc.

What do we want to list?
The Idea of the last try was "RAW DATA", I don't like this personally, people had to calculate frame advantage and the stuff by themselves and this encounters problems, people just don't want to do stuff themselves, they want to open something and read what they want to know. (totally understandable)

How do we share frame data?
we did the last one through google docs. Was this good? Were we satisfied with that? Did it work?
Better ideas?
I still have the google docs thing in my bookmarks but it didn't seem to be that active overall, don't know why :/


Before we really start this project all over again, I want you to ask this questions because it's a "community" project and everyone can add his opinion, share ideas, whatever.

the aim is to create the ultimate frame data base of this game!

looking forward to it

 

[MonkUnit]

I think Damtarion has a web page with all the frame data for character hitboxes.

 

[KayLo!]

^ Dant's site is decent but definitely not polished enough imo.

Sorry for upcoming tl;dr.....

--

I'm with you, Yika.... I didn't like how the last attempt ended either (although tbh, I think it's one of those massive projects that has very little chance of ever getting 100% done.....). I don't have a Wii atm to do any more testing, but since this is the one thing I'd like to see completed, I'll help however I can!

As for your questions....

For the final release form, I'm torn between thread, Excel (GDocs), or an external web site. Threads are the most obvious answer, but personally I don't think threads allow for fluid enough formatting, and with all the info that's eventually going to be compiled, it's bound to look cluttered. There's no way around it. An Excel sheet or well-formatted web site would make organizing the data much easier (not to mention it'd be easier for us to make edits in the future)..... so ultimately, I think we should go with one of those and make sure they're linked to in each character forum's frame data/research thread. This will also allow each forum to keep doing additional data however they want without us saying "hey, you have to follow our format for your official data" or creating duplicate threads.

How do we share data? GDocs was good imo. It's the easiest way I can think of to have a central hub of info that we can freely edit & share.

The hardest question is exactly what data we want on the list for the final public version, and I think that was one of the major problems with the last attempt at this. Nobody (myself included when trying to come up with formatting ideas) really knew exactly what was being included, so there was raw data flying around everywhere -- some characters had holes in their basic data while things like knockback calculations and frame advantages for specific percents were being done. Made no sense.

Imo, this is a rough idea of what should be released for each move (off the top of my head, so it might be missing stuff):

- Basic duration info -- startup, hitbox out, FAF type stuff
- Charge release or whatever you call it for smashes
- Landing lag, RCO lag, and autocancel frames for aerials
- Fresh or base damage
- Adv on block (only for ground moves I guess unless we can find an easy way to test & list it for aerials)
- Invincibility/super armor/absorption/reflect frames if applicable
- Trip rate if applicable
- SDI/hitlag multipliers
- Length for rolls
- Launch angle?
- Collision type?
- Element?
- Specials will obviously be on a case-by-case basis since they're all extremely different​

Now, I'm not saying we shouldn't find/calculate more detailed stuff for ourselves or for any character-specific research threads or whatever, but for the basic, official frame data release, I don't think overloading it with crap most people aren't gonna remember/really care about is a good idea. I feel like less is more for a project like this, and if we want to add stuff later, we can. But let's at least get the most basic **** out there for now.

 

[Yikarur]

thanks for your input, I really appreciate it

I don't like the use of FAF in frame data though, I prefer "Duration" or "Ends" (duration is better)

 

[KayLo!]

Oh, whichever we end up using, doesn't matter. I prefer FAF, but as far as my last post, I just meant something like that should be included. =X

 

[Yikarur]

so it'l by like

| Start-up | Hitbox | Duration |Base Damage | Hitlag | Shieldstun | SDI Multiplier/ Hitlag Multiplier |Angle |
Move |

whatever.

going to be a big table xD

we also could use a wikipage. "Brawl Frame Data Wiki"
I'd like this idea xD
it's super free editable and it could contain everythiiiiing xD

 

[MonkUnit]

Trying not to get into this "FAF vs IASA" debate, but I would suggest using IASA as it is more of a common term for smash than FAF is.

 

[Yikarur]

IASA is wrong for almost everything. (you don't interupt anything)
we should use "Duration", the melee frame data uses that as well (+IASA where it's applicable but Brawl IASA is stupid)

 

[KayLo!]

Oh yeah, I knew I'd forget something. SDI and hitlag multipliers should definitely go in there too.

 

[MonkUnit]

IASA is wrong for almost everything. (you don't interupt anything)
we should use "Duration", the melee frame data uses that as well (+IASA where it's applicable but Brawl IASA is stupid)

Sounds good.

 

[KayLo!]

So when we list duration, are we saying it's at the animation's end or at the point when the next frame is FAF? I'm guessing the latter, just clarifying.

 

[Yikarur]

I'm for the latter.

 

[Dr. Tuen]

I think we need to break this into stages. If we go all gung ho like this, it may end up like the last project. If we instead take things one step at a time and release information out the public at regular intervals, we'll up our reputation by quite a bit and keep our focus.

That said, we could go by character (full data for each) or we could go by data type (start with start up, end/IASA/FAF, duration, minimum/maximum charge frames [variable charge projectiles, smashes], hitframes; move on to add hit/shield stun, multipliers and launch angles... something like that).

In either case, we should also look at what information we already have and are willing to recycle for this project. Information on the boards can be helpful too, if validated.

We also want to address how we want to communicate this information amongst ourselves and how we want to present it to the boards. I've been doing some work with ZSS on frame traps and I utilized their frame data spreadsheet... and it's not the easiest thing to interpret.

===

If it were my choice, I'd say we go character-by-character. That way we show a steady stream of information release and have steady focus for the project. It'd be much easier to cover previous information and validation that way too.

===

Also, I think we should consider adding things like dash/walk animation frames (start up, stop, and possibly fox-trot length/frames) along with jumping frames (start up, the frame that determines short/high hop, the dead frame, etc). For characters that glide, I'd like to see glide canceling frames as well (I'm looking at no ground contact, with a jump remaining so you don't freefall).

 

[Yikarur]

Jump Stuff belongs in his own section but we will cover that one for sure.

what do you people think about my wiki idea?

 

[Dr. Tuen]

Jump Stuff belongs in his own section but we will cover that one for sure.

what do you people think about my wiki idea?

I think this is a fantastic idea. "SmashLabs Wiki" or something, heh. If we maintained a thread in SWF that posted about updates to the wiki, that would serve as our PR thread with the rest of the boards.

It would also allow us to link various concepts without having to repeatedly explain them. It would result in statements to the effect of "X character's usmash has a Y hitstun multiplier" with the underlined linking to a page which explains the concept and the appropriate calculations.

Obviously we'd end up formatting it a lot nicer, but that's the effect.

 

[Yikarur]

thats infinitely awesome :3
we could use templates and stuff for frame data tables.

 

[MonkUnit]

There's no such thing as a "hitstun multiplier"..

There's hitlag multiplier though.

 

[rPSIvysaur]

what do you people think about my wiki idea?

Do want.

Someone make it happen.

 

[KayLo!]

+support for Wiki

 

[phi1ny3]

I like this idea

btw, I'm always a little sad that all people ever look at is startup time. Yes, it is really important, but at the same time, they don't see all the tidbits (I always get dumb theorycraft over "Lucario so slow, you lose", and it's like "A: It's not terribly slow to the point where he's like low-tier for frame startup, and at least it's enough for aura and range to compensate, and second, have you seen how long some of his hitboxes last, or how little cooldown some attacks have?" lol), it sorta feels like I'm unraveling my character's own subjective demise lol, but hey, someone's gotta do it.

 

[Stealth Raptor]

i really like the wiki idea, so that stuff can be interlinked and definitions are easy to come by while looking at hard data. honestly, if all of us do like, five moves a day, we could be done with the entire cast within a few weeks. i thikn google docs can be used to compile data as we test and double check, then we can transfer the data over to the wiki. have a couple people working on a standard of how its gonna be layed out, standard list of definitions and so forth, and the rest just work on frame data. if we really put outselves into it we could have the data compiled easily in less then a month, and have the full wiki up within two

 

[Yikarur]

do we have ANYONE who is good in stuff like designing wiki etc?

I'm not the man for it xD

I can transfer old templates I used earlier in the German Smash wiki (it never got active though).

 

[Stealth Raptor]

do we have ANYONE who is good in stuff like designing wiki etc?

I'm not the woman for it xD

I can transfer old templates I used earlier in the German Smash wiki (it never got active though).

fixed

i vote for kaylo, shes good at making things look pretty

 

[Yikarur]

Installing a Wiki is not a easy job, has anyone experience with that?

 

[KayLo!]

I think we need to break this into stages. If we go all gung ho like this, it may end up like the last project. If we instead take things one step at a time and release information out the public at regular intervals, we'll up our reputation by quite a bit and keep our focus.

...

If it were my choice, I'd say we go character-by-character. That way we show a steady stream of information release and have steady focus for the project. It'd be much easier to cover previous information and validation that way too.

Agree with the first part but disagree with character-by-character. Imo we should break it up by data type, starting with the bare bones basics. Maybe something like this:

Phase 1: Basic duration info (startup, hitbox out, duration/end/whatever), damage, landing lag/RCO & autocancel frames, charge release, invincibility/super armor/absorption/reflect frames
Phase 2: Trip rates, block adv, SDI/hitlag multipliers
Phase 3: Collision types, launch angles, roll lengths, elements, everything else -- basically **** most people don't really care about​

My reasoning is that if for some reason we don't finish this (which, let's be realistic, might happen), it would look dumb for a handful of characters to be done and not others. I think it's much better for every character to have the basics at least -- which honestly should have been done ages ago -- than for some characters to have everything while others have nothing at all.

do we have ANYONE who is good in stuff like designing wiki etc?

I've never worked on a wiki before, but I've done basic HTML/CSS sites, and I have a good eye for formatting.... so I can give it a try. Preferably with the help of someone who's had experience with editing wikis. (I'm guessing it can't be that hard tho?)

 

[rPSIvysaur]

Does Toomai know? I'm pretty sure he has some wiki experience.

 

[Toomai]

Gimme a few hours to muck about.

RESULT: Alright after severe thrashing (mostly digging up an Apache installation that seemingly atrophied since last year) I have successfully put a wiki on this computer. Combined with this I think we can get the wiki idea to work.

First we'd need to bug Bio until he gives us a place to put stuff. Either I need generally-unrestricted access to that location or he has to deal with all of this. Once that's done I'll/we'll need a bit of time to get things initialized. Set up boring help pages, determine naming conventions, make some templates, create non-default skin possibly. After that we can start inserting data and see how it goes from there.
I nominate myself as biggest cheese of the wiki due to my obvious experience and general shenanigans etc.
If "Smash Lab Wiki" isn't good enough for you we could do something spiffier like "Smash Bros. Knowledge Node", "Smash Lab Infobase", or even "Smashic Serendipity Centre".

 

[standardtoaster]

 

[KayLo!]

=X!

K, this wiki stuff is waaay more complicated than I thought, so.... y'all got that. But I'll definitely help with the pre-wiki info collection.

 

[Dr. Tuen]

=X!

K, this wiki stuff is waaay more complicated than I thought, so.... y'all got that. But I'll definitely help with the pre-wiki info collection.

I'll second this. I'm not really that great with web related things :-/. I like the initiative though!

 

[DeLux]

IMO if Toomai maintains it, he gets naming rights. But Yika is Directing this, so he can decide

 

[Yikarur]

fear my power!

 

[MonkUnit]

fear my power!

uhhh n o

 

[Isatis]

Toomai, don't you own & run SSBWiki? (and have all the upload rights and such?) If you do I have a better idea.

 

[Toomai]

Toomai, don't you own & run SSBWiki? (and have all the upload rights and such?) If you do I have a better idea.

I'm SmashWiki's chief guy but I don't technically own or run it; Porplemontage does. (He hosts SmashWiki as well as five other wikis (including Super Mario Wiki).

What were you thinking? Branch SmashWiki or something? (For what it's worth I was going to suggest affiliation at least once we got things going.)

 

[Isatis]

Ah. One idea I had was making a subdomain (smashlab.ssbwiki.com) and having another Wiki put into that.

I can set up MediaWiki really quick on a spare server I have though.

 

[Yikarur]

Beware that we probably don't want to let Non-Labbys Edit anything.
The last thing I want is false information spreaded there

 

[Toomai]

Beware that we probably don't want to let Non-Labbys Edit anything.
The last thing I want is false information spreaded there

Well, it's easy to stop non-logged-in people from editing. Restricting the generic user group might be a bit more challenging, since wikis are generally designed to allow more people to edit as opposed to fewer. I'm sure some usergroup fiddling can get it done - though that would also stop people from posting on talk pages (I think locking namespaces for usergroups is very unwikilike), which would probably be preferable to posting in the link thread here for those who want to do something like "Don't get X" or "Number Z looks wrong".

 

[Stealth Raptor]

one thing that helps from data standpoint is we have chats like pika with full frame data by lab people, so we should have some of the work out of the way. once we have a standard of what exactly we are going to put on each page we can start data collection. i have a large chunk of samus data done personally

 

[Stealth Raptor]

so any progress on this?