Meta Knight Panel (Topics for Testing)

[DeLux]

Please Post Things in here regarding the MetaKnight Panel. Please do not spam or I will be angry.

Topic - Meeting Times

TO DO:
Research Topics

 

[DeLux]

Double post to reserve in case I need it.


I'm free to meet late tomorrow night (Central time), plus late thursday night as well. Friday is my birthday, so I'll be going hard all weekend I assume.

 

[rPSIvysaur]

I'm cool most of the time as long as it's after 3PM (on weekdays)

 

[Player-4]

I think a thread for talking would be better than Skype, personally.

 

[Yikarur]

this panel stuff is so overrated D: even an own subsection.. really not necessary imo.

 

[Orion*]

Today and tomorrow I will be awake... so i can skype. However I don't know how sober I will be >_>

Today I can easily make an excuse to only have a few beers since I have class tomorrow so that would be better for me. past 6am GMT+2 (amsterdam time *****s we out) then I doubt I will be happy to talk though.

This weekend Im going hard probably to, my friend has her bday and shes irish so drink is going everywhere lolololol.

So yeah today or next week is best for me imo

 

[DeLux]

We could probably end up doing both tbh. Worst case scenario, Orion you can just post some stuff on your own time. I figured the time difference might bring some issues, so that's fine.


As long as we're active and moving forward, I don't care where/how the discussion takes place.

 

[Exdeath]

If I have free time and Internet access, I'm probably on Skype.

My main "free" days are on Friday and Sunday.

 

[Sensei Seibrik]

i should be available MOST nights around 10-11pm IF NEEDED

 

[Yikarur]

Sebrik you're always in this "if needed" mood like "I actually don't care but if it reaaally needs to be done then ok"

 

[KayLo!]

I was thinking that lol. Why join the panel if you don't want to actively do the work? x.o

 

[Sensei Seibrik]

i was invited, i didn't apply for the position. if you need me to do something, i'll do it, simple as that.


i'm a tool, don't ignore me cuz im not "excited", use me as you need me

 

[DeLux]

lol, everyone relax. why don't we work something for tomorrow night? On skype preferably but we could do it here with the new autoupdate SWF.

In the mean time, start thinking MK metagame.
If anyone has any suggestions for discussion we can start that now.

Otherwise I'll direct it towards just basic frame data stuff you'd like us to test. Looking for "guaranteed" combos, frame traps, etc.

 

[Sensei Seibrik]

well as far as punishing mk, it's pretty understood that one needs to predict the mk's move in order to punish it directly. (this is unlike most characters u may be able to punish on reaction)

IMO, that makes for a GOOD "best character" cause it requires something more than tech skill to defeat, it takes brainssss

 

[DeLux]

We're just trying to push the tech limits and knowledge on the character closer to the theoretic maxims that make him the best. And trying to operate outside of the parameters of a player.

The discussion shouldn't be necessarily geared towards beating MK. It should be geared towards knowing MK. If that makes sense at all.

Confirm on a time tomorrow night? Seibrik is already listed at 10pm central as most likely available. rPSI set for after 3 so again 10pm works.

10pm would work for me if I skipped some plans. Unless it's deemed I'm not need.

 

[Exdeath]

I would like to start with Side-B canceling on Rainbow Cruise, Norfair and in general.

 

[DeLux]

Alright, we have ealert working on the that project officially speaking.

 

[DeLux]

Is the current MK frame data thread acceptable to leave alone?

Also additional testing projects:

U-Air String Evaluation: At what percent for each character does U-Air stop stringing based on knockback/frame testing

Planking: Conclusive frame testing based to incorporate element of spacing into tests outside of only a frame examination

Shuttle Loop Mapping: Spacing distances that determine how exactly shuttle loop behaves when wrapped around a platform or edge. Glide momentum loss rules.

A couple of my ideas to look into.

 

[Exdeath]

The MK frame data needs to be re-done. Ftilt/glide attack in particular. The rest that you listed is good, too, but in particular it would be nice to know the game re-classifies glide landing(s).

 

[rPSIvysaur]

The rest that you listed is good, too, but in particular it would be nice to know the game re-classifies glide landing(s).

I'll be on this one.

 

[DeLux]

I'll start looking for someone to do the MK frame data. If they spread sheet it, I can make it all nice and pretty in the template we were using.

 

[rPSIvysaur]

I can spreadsheet, I just hate formatting it. lmao

 

[DeLux]

But,

The Gimmicksoogle Wave...

 

[rPSIvysaur]

I'd rather spreadsheet than Gimmick Wave.

 

[KayLo!]

I can spreadsheet, I just hate formatting it. lmao

I already told you I'd do any formatting type stuff.....

Actually, now that we have tables, I've been semi-working on a new format for the Pika/Zelda data.

 

[DeLux]

Legit, so I'll leave data collecting to rPSI, formatting to KayLo

Can we shoot for a deadline in the next 10 days for the current projects?

 

[Yikarur]

U-Air String Evaluation: At what percent for each character does U-Air stop stringing based on knockback/frame testing

way too much work for less "reward". You know this, no one is going to read and learn that

 

[Exdeath]

way too much work for less "reward". You know this, no one is going to read and learn that

I would and I'm sure that any Meta Knight who puts effort into studying Brawl would (a.k.a. Orion/Havok/etc. probably would).

 

[rPSIvysaur]

The problem is the number of permutations it would take to actually gather the "reward"

Here's the variables:

1) Characters - 37
2) Percent - let's say 0-40
3) Staleness - 10!
4) Double Jump Height - 5 (since each jump will put you in a different position)

That's an absurd amount of testing to do for something that is just done by reading DI normally, and also doesn't take into account SDI, which makes even harder to test. It'd be cool to research and get down, but I don't want to allocate Smash Lab resources to it until we have major info released on every character. If anyone wants to take it up as independent research, they can go ahead and do that, but I'm not going to require anyone to test it.

 

[Exdeath]

Specific frame data to check is Dtilt. In particular, meaty Dtilt (how to get it/what it looks like/etc.).

 

[Toomai]

D-tilt is whack. Its four hitboxes all deal different damage (the weakest being the farthest away). Weirder still, the two stronger hitboxes make a weaker sound on impact.

Also, it hits on frames 3-4 and can be interrupted frame 16.

3) Staleness - 10!

Technically it's 2^9, not 10!.

 

[rPSIvysaur]

herpa-derpa bad at maf moment, lol. It's been forever since I've done probability/stats stuff.

 

[Exdeath]

D-tilt is whack. Its four hitboxes all deal different damage (the weakest being the farthest away). Weirder still, the two stronger hitboxes make a weaker sound on impact.

Also, it hits on frames 3-4 and can be interrupted frame 16.Technically it's 2^9, not 10!.

I didn't realize that Dtilt had quite that much variation, but my main point was toward safety. I believe that Dtilt might not be shield-grabbable for some of the cast. I don't mean spacing it to poke; I mean that it might be useful as a boxing tool.

 

[DeLux]

Dnyce is on the dtilt project. You should talk to him in skype chat to coordinate exactly what you want.

At the moment, we were looking into combo advantages on successful hit, not necessarily hitting an opponent's shield.

 

[Exdeath]

Dnyce is on the dtilt project. You should talk to him in skype chat to coordinate exactly what you want.

At the moment, we were looking into combo advantages on successful hit, not necessarily hitting an opponent's shield.

Hit is more important, however I was just thinking of something in particular. I'll get with Fino when I can haha.

 

[rPSIvysaur]

Wanted to update you, I have the comprehensive moveset data of all regular attacks done. I'll probably be translating the flags soon enough for even more comprehensiveness :D. I still have random various stuff such as specials, ledge stuff, get-up stuff, glide, etc. to do, but it's coming along.

 

[Exdeath]

To specify on ledge-related get-ups, can you see if there's an optimum way to hop onto the stage? The more traditional (a.k.a. M2K started it) is to ledge drop>Uair>Jump>Air Dodge, but Seibrik as told me before that it's faster to not air dodge.

 

[Sensei Seibrik]

well, u have less landing lag cuz airdodge has more landing lag than normal landing i believe

so u can just hold the stick at an angle as u jump toward the stage (down and toward the stage) so u dont grab ledge but u like, glide->walk onto the stage, u can do it into any input (dtilt, dash grab, attack, roll, upb) w.e.

m2k likes airdodge (i think) cuz he can buffer durin the airdodge to use his input on the first possible frame, as appose to relying on his own reaction/timing

also the airdodge can be used to get passed some smash attacks, mixup between using airdodge and not and it should be really hard to time a hit on u from the ledge

 

[rPSIvysaur]

Air dodge goes into the same landing action as regular landing. The only difference made would have to be a hard landing vs. a soft landing, which wouldn't change if you're doing an airdodge or not.

As far as progress goes for right now:
I have all of the frame data and hitbox data compiled for MK except specials (still waiting to hear back from Shanus about Nado) and KayLo is currently formatting said data for neatness/release. If anyone wants the spreadsheet, I'll send it to you. Fino is currently working on getting D-tilt hit advantages and EA (I think it's EA) is working on Side-b cancel shenanigans. Keep up the work guys.

 

[DeLux]

Yeah, we're I'm hoping for us to get STL 2.0 out this weekend. And then maybe move forward with releasing all this stuff shortly after. So there's time, but hopefully KayLo, Fino, and EA get there stuff in soon

I think mainly KayLo formatting is the main one. We could insert Fino and EA's stuff post release if we wanted to.