Complete Hitbox/Frame Data For Every Character (Now with KO Percents)

[Thinkaman]

Paltuna's tilts beat all spot-dodges and are disproportionately good against shields. However, they are harshly punished by rolls.

 

[dragontamer]

Somewhat. Meaty attacks with large hitboxes (like ftilt or up-tilt) seem like they'd be hard to roll around. I'll have to play with her in practice to see.

The main issue I've had with her is that her approach game is awful. I'd also have to come in with a gameplan to force opponents to block.

 

[Thinkaman]

Somewhat. Meaty attacks with large hitboxes (like ftilt or up-tilt) seem like they'd be hard to roll around. I'll have to play with her in practice to see.

U-tilt can beat the closest of rolls (closest on ending), but why would you do a u-tilt as an answer to a roll? At least do a d-smash, if not an f-smash. U-tilt's value proposition in ground-to-ground combat otherwise isn't very high.

The main issue I've had with her is that her approach game is awful.

Do you have a moment to talk about our lord and savior Super Speed?

 

[dragontamer]

U-tilt can beat the closest of rolls (closest on ending), but why would you do a u-tilt as an answer to a roll? At least do a d-smash, if not an f-smash. U-tilt's value proposition in ground-to-ground combat otherwise isn't very high.

Less of an "answer", and more of a "Holy crap, 31-active frames is meatier than some people's entire roll". Active from frame 10 to 41. If I'm expecting a roll, of course Down-Smash is the go-to answer.

I guess if the opponent rolls into an attack with lol 31-active frames, that was their fault, as opposed to me outplaying them. And I just tested up-tilt, it doesn't seem to be a very good ground based attack as you said.

Do you have a moment to talk about our lord and savior Super Speed?

Not while I'm still arguing with you in that other thread :-)

 

[Nobie]

I don't know how you'd express this in the chart, but one thing I've noticed about Mega Man is that while Metal Blade doesn't KO when shot directly, it can KO if thrown as an item.

 

[Thinkaman]

I don't know how you'd express this in the chart, but one thing I've noticed about Mega Man is that while Metal Blade doesn't KO when shot directly, it can KO if thrown as an item.

Grabbable items are actually universal entities that are not tied to particular characters from a programming standpoint, and are not in these files at all.

 

[Unknownkid]

Can you explain how you are doing the shield bonus analysis?

 

[dragontamer]

Can you explain how you are doing the shield bonus analysis?

Mastercore2 script dumps.

One of the parameters to hitboxes seems to be the shield bonus. Thinkaman seems to have some scripts to automatically parse the data and perform calculations on it.

 

[ChronoPenguin]

How much health do shields have in this game?
Is it known? Im guessing 50?

 

[Thinkaman]

Can you explain how you are doing the shield bonus analysis?

Grabbing the raw data from the extracted game files (3DS) dantarion provided.

How much health do shields have in this game?
Is it known? Im guessing 50?

Yup.

 

[Indigo Jeans]

So Brawl's KB formula CAN be applied to Smash 4 with more or less accurate results? If so feels good

This also means I should be able to extrapolate gravity values

When figuring out whether Brawl's formula fit, how were you doing it? I had been thinking, of, say

Testing for Vlager NAir strong hit KB
-Villager@0%
-On Lucario@0%
-hit in match, record Max Launcher Speed value

Ive heard that Smash 3DS's MLS value is accurate to "real" knockback, as opposed to Brawl's that had to be divided or multiplied by 0.33 or something, I don't remember honestly

Lucario's weight is known, NAir has horizontal KB, meaning gravity can be ignored for KB calcs, rage can be ignored, and the freshness bonus of x1.05 can be accounted for. Using this data I could simply run a calculation for what the KB should be using Brawl's formula, then compare it against the MLS value

I didnt make myself terribly clear, but that was the process I was thinking of using to check if the formula was still applicable. Do you do somethig like that, or...?

 

[Unknownkid]

Grabbing the raw data from the extracted game files (3DS) dantarion provided.

Oh, I see. Apologies for the stupid question. I should assume it is still in development since characters like Little Mac, MetaKnight and Lizardon information are missing while the full data (http://pastebin.com/PzTjdtx8) show they have shield bonus potential.

 

[Indigo Jeans]

That sort of info is simply unparsed in mastercore at the moment. It's there, but it's not labeled.

Edit: ran the calc for standard tree's strong falling hit.

Calculated value = 100.17022613
Max Launcher Speed value = 63

Hmmm

 

[Unknownkid]

I understand now. Please excuse my naive comments.

 

[Nysyr]

-snip-

Are you taking into account aura in terms of knockback? The Knockback listed doesn't seem to line up with a 0% Lucario, nor a max Aura Lucario.

 

[Indigo Jeans]

Lucario's aura doesnt affect the KB of other players' moves, does it?

I understand now. Please excuse my naive comments.

Hey, it was an honest mistake, no need to put yourself down over it!

 

[Nysyr]

Lucario's aura doesnt affect the KB of other players' moves, does it?


Hey, it was an honest mistake, no need to put yourself down over it!

Sorry, misread.

 

[Thinkaman]

That sort of info is simply unparsed in mastercore at the moment. It's there, but it's not labeled.

Edit: ran the calc for standard tree's strong falling hit.

Calculated value = 100.17022613
Max Launcher Speed value = 63

Hmmm

Max launcher speed is a useless statistic that has poor (if linear) correlation with actual knockback.

It is scaled funny, and rounded down to the nearest multiple of 13.3333.

It's some sort of unit conversion behavior.

Edit:

Wynaut? Biggest Hitbox Size Table

This is a silly chart that is SUPER misleading, because it in no way accounts for multiple hitboxes, how much they do or not not overlap, nor how disjointed this biggest hitbox is. It's JUST raw radius of the single biggest hitbox, rounded down.

For reference, I think the biggest standard moveset hitbox in the game is the outer hit of Volatile Breathing, which is size 48.0.

Edit2: Added grabs.

Edit3: Multiplied all the values in the size table by 10, for better comparisons. If we're going to have a table of mostly useless numbers, lets at least have them be slightly different, and not all rounded down to 3 or 4.

Edit 4:

Biggest Standing and Pivot Grab Hitboxes (not quite the same as range, but close)
koopa 6.0
donkey 5.8
yoshi 5.5
lizardon 5.4
murabito 5.0
dedede 4.5

Biggest Dash Grab Hitboxes
yoshi 5.5
koopa 4.8
donkey 4.6
lizardon 4.3

 

[ParanoidDrone]

That hitbox size table goes exactly nowhere. The URL looks like this when I click on it.

http://biggest%20hitbox%20size%20table/

 

[Unknownkid]

I believe that's the joke...?

 

[dragontamer]

Wynaut? Biggest Hitbox Size Table

"URL='http ://Biggest Hitbox Size Table'"

I think you forgot a link there or something.

 

[san.]

Hitbox size table: http://pastebin.com/vRk1UsHe

Looking over it briefly, it's not indicative of too much, as explained in Thinkaman's post.

 

[Thinkaman]

Weird, sorry for the broken link. Fixed.

 

[Indigo Jeans]

Max launcher speed is a useless statistic that has poor (if linear) correlation with actual knockback.

It is scaled funny, and rounded down to the nearest multiple of 13.3333.

It's some sort of unit conversion behavior.

Hm... so then what is the relationship between MLS and KB? This is really odd considering the Wii U version gives a totally different value for MLS. On the 3DS the units are dimensionless, while on the Wii U it gives miles per hour. And things like Villager's jab1 have a non-zero MLS on the 3DS, but the Wii U version gives 0 mph

 

[Thinkaman]

Hm... so then what is the relationship between MLS and KB? This is really odd considering the Wii U version gives a totally different value for MLS. On the 3DS the units are dimensionless, while on the Wii U it gives miles per hour. And things like Villager's jab1 have a non-zero MLS on the 3DS, but the Wii U version gives 0 mph

My solution is to disregard both completely.

 

[Pyr]

To bump, and to ask a question: Is there really no difference in the 3 different directions for Ganon's FSmash? Up, down, and neutral are all exactly the same.

1. F-smash (high)
2. Frame 21-26: 24% 35b/83g (KO@ 71%) 40° Darkness
3. Frame 21-26: 24% 30b/75g (KO@ 85%) 40° Darkness
4. Max Damage: 24%
5. 
   
6. F-smash (normal)
7. Frame 21-26: 24% 35b/83g (KO@ 71%) 40° Darkness
8. Frame 21-26: 24% 30b/75g (KO@ 85%) 40° Darkness
9. Max Damage: 24%
10. Smash charge window on real frame 10
11. 
    
12. F-smash (low)
13. Frame 21-26: 24% 35b/83g (KO@ 71%) 40° Darkness
14. Frame 21-26: 24% 30b/75g (KO@ 85%) 40° Darkness
15. Max Damage: 24%

Seems silly.

 

[Lavani]

There really isn't.

I'm actually surprised by that, I thought angling it did make a difference in angle/power for him, but I guess not.

 

[B!squick]

Okay, I probably should have mentioned this when I first noticed so I'm sorry for not doing so sooner, but you have Koopa's aerial SideB switched with his ground SideB. So sayith Hawaiians and the fact that the aerial version certainly doesn't appear to be twice as fast.

EDIT: Also, for your KO percents you say you calculated it based on a certain angle. Probably a stupid question but did you account for moves that hit straight up or to the side like Purin's Rest?

 

[Nu~]

Wait...pac-man's fsmash kills earlier than his usmash? I tested it on training mode and up smash always killed earlier, unless some vertical vectoring is still present in the game?

 

[Kofu]

Wait...pac-man's fsmash kills earlier than his usmash? I tested it on training mode and up smash always killed earlier, unless some vertical vectoring is still present in the game?

Who did you test it against? Some characters are more resistant to vertical KOs. Thinkaman just tested KO percents against a Mii, and I believe Miis don't have particularly good vertical resistance.

 

[Nu~]

Who did you test it against? Some characters are more resistant to vertical KOs. Thinkaman just tested KO percents against a Mii, and I believe Miis don't have particularly good vertical resistance.

I tested it against a Zelda, so that may be why. Just for the future, what type of characters In general are more vulnerable to horizontal KB as opposed to vertical and vice versa?
Does being floaty or falling quickly have to do with it?

 

[Kofu]

I tested it against a Zelda, so that may be why. Just for the future, what type of characters In general are more vulnerable to horizontal KB as opposed to vertical and vice versa?
Does being floaty or falling quickly have to do with it?

In general, characters that fall fast are harder to kill vertically and those with fast air speed are harder to kill horizontally. I don't know all the details, though. Try using his smashes against Dedede. I can guarantee you FSmash will kill earlier by a considerable margin.

 

[DeLux]

ETA on when we'll have full frame data featuring FAF/IASA's?

 

[Thinkaman]

ETA on when we'll have full frame data featuring FAF/IASA's?

This data seems to be in animation files this game. We need someone to crack them, and I know next to nothing about animation files.

 

[Kofu]

Would it be much of a hassle to re-add the substate reference numbers for moves, or at least specials? It makes it easier to go back and reference more of the move data in Dantarion's files.

 

[Meru.]

Some of the KO percents seem a bit off to me... Peach's dash attack killing at 186%? I often kill at like 130% with it. Is it perhaps because of Rage? Dash Attack has quite a lot of knockback growth since it only does 6% so perhaps the higher the knockback growth the bigger the Rage effect?

I also could have sworn Donkey Kong's Bthrow was stronger than most other Bthrows... weird @_@

 

[David Viran]

For zero suit the kill percent said up b killed at 130%. How was this measured because I was testing it and it was killing at 113% from the center of FD with the best DI.

 

[dragontamer]

For zero suit the kill percent said up b killed at 130%. How was this measured because I was testing it and it was killing at 113% from the center of FD with the best DI.

The important thing is that the values were automatically generated from a script through calculations and statistics based on the 3DS version.

The 3DS has larger blast zones. So on the WiiU, the percentages will be a bit of an over-estimation.

That said, Thinkman can probably rerun his script based on some WiiU values... because this criticism seems to come up a lot. I'd expect the WiiU version to be the "primary" game. Smaller-blast zones will also emphisize BKB a lot more than right now (making Luma's utilt a fair bit better on WiiU than 3DS)

 

[David Viran]

The important thing is that the values were automatically generated from a script through calculations and statistics based on the 3DS version.

The 3DS has larger blast zones. So on the WiiU, the percentages will be a bit of an over-estimation.

That said, Thinkman can probably rerun his script based on some WiiU values... because this criticism seems to come up a lot. I'd expect the WiiU version to be the "primary" game. Smaller-blast zones will also emphisize BKB a lot more than right now (making Luma's utilt a fair bit better on WiiU than 3DS)

I wasn't trying to critic him I was asking how he got the data. Thank you for your reply.

 

[Indigo Jeans]

Some of the KO percents seem a bit off to me... Peach's dash attack killing at 186%? I often kill at like 130% with it. Is it perhaps because of Rage? Dash Attack has quite a lot of knockback growth since it only does 6% so perhaps the higher the knockback growth the bigger the Rage effect?

I also could have sworn Donkey Kong's Bthrow was stronger than most other Bthrows... weird @_@

Rage doesnt even matter. High kbg moves KO early because, well, they get stronger faster. The amount of kb a, say, 105 kbg move applies will increase, say, 2 times as quickly as that of a 90 kbg move, even if the second one has higher bkb.

Rage isnt responsible for differences in KO percents in a comparison like that