Character Height: A Rough Start

[Dr. Tuen]

INTRODUCTION​

The purpose of this thread is to release and review preliminary character height data. These finding include character standing heights and character crouching heights. This is useful for determining who can be struck by falling aerials, rising aerials, and various grounded moves. While most important for the shortest characters in the game (Kirby, Pikachu, etc), some may find other application for this data.

While this will be discussed further, it is worth saying here: This is very rough data. It was based on observation. There are many sources of error.

Methodology​

A custom stage was created in order to have a wall to measure height by. That wall had bricks, and character height was estimated based on their position with respect to those bricks. Those estimations were used to create a preliminary height order. Any pair of characters that had matching or near matching height were compared directly and re-ordered if necessary.

Spoiler: Height Measurement Custom Stage

Most characters have animation cycles that adjust their height. For those characters, low and high heights were recorded and an average was taken. If that average tied with another character, the character with the higher variance was considered to be "taller."

Some characters had significant features that required characterizing two heights. For example, when King Dedede crouches, his head dips far lower than his hand. His crouching height is characterized twice for that reason.

This method for assessing character height is riddled with error. Heights were visually assessed with a ruler taken to my screen, against a custom stage, while I had to pause to catch them in different parts of their standing/crouching animations. I also do not know where the hitbox for each character ends, so I tried to estimate the "top" of each character.

I encourage discussion on how to gather more accurate height data. This is all preliminary, and it needs revision.

RESULTS​

As a reminder, all heights are in "bricks," and some heights have higher significant digits due to it being the average of two representative animation cycle heights.

Spoiler: Character Height Ranking

 

[Cassio]

Nice, I'm glad someone did this. I had a personal (far less extensive ) version of this, seems similar to what I had. This may be a good thread for the academy forum.

 

[ぱみゅ]

Can walls be put on top of each other (as in, to make a taller wall)? If so, isn't it a bit easier to identify heights using Giant Smash?

Anyway, moving to the Academy.

 

[Dr. Tuen]

Can walls be put on top of each other (as in, to make a taller wall)? If so, isn't it a bit easier to identify heights using Giant Smash?

Anyway, moving to the Academy.

I think a taller wall just takes a longer line drawn in the stage editor. Giant smash could be good! Thoug. The pause camera's closest zoom distance gets pushed out for bigger characters. If that applies to giant characters, it may introduce a new kind of variation to the assessment process. Still, worth a look. Thanks!

And thanks for the thread move, heh.

 

[Indigo Jeans]

Wii Fit Trainers ass, lmao

Good stuff Tuen

 

[Ffamran]

I wonder how this translates to actual (centi)meters and/or feet. Falco's 6' 2" or 1.88 m, Fox is 5' 7" or 1.73 m which is the same height as Little Mac's current, updated height, Samus is 6' 3" or 1.9 m, but I don't know if that's her height in her armor or not, and Ganondorf's supposedly 7' 6.5" or 2.30 m in OoT according to Hyrule Historia. Well, considering that the Pits are taller than Falco and they're technically 13 despite being angels... Man, they're tall teens. And... Bowser being 7 ft something isn't farfetched considering he has grown and shrunk from being the size of a small planet to fitting in a large go-kart fine.

 

[Dr. Tuen]

Wii Fit Trainers ***, lmao

Good stuff Tuen

Hilarious, and relevant for projectiles (ha). Though I think she can crouch under all of them...

 

[Ansou]

Wouldn't it be a lot better to gather this data using a capture card? Then you can just have a fixed camera and measure whichever frame you want in pixels without pausing which seems much more accurate. Now I'm guessing you don't have a capture card, so that's obviously a problem. I could probably borrow my friend's capture card to record all stand and crouch animations in 1080p 60fps if that would be helpful.

 

[erico9001]

INTRODUCTION​

The purpose of this thread is to release and review preliminary character height data. These finding include character standing heights and character crouching heights. This is useful for determining who can be struck by falling aerials, rising aerials, and various grounded moves. While most important for the shortest characters in the game (Kirby, Pikachu, etc), some may find other application for this data.

While this will be discussed further, it is worth saying here: This is very rough data. It was based on observation. There are many sources of error.

Methodology​

A custom stage was created in order to have a wall to measure height by. That wall had bricks, and character height was estimated based on their position with respect to those bricks. Those estimations were used to create a preliminary height order. Any pair of characters that had matching or near matching height were compared directly and re-ordered if necessary.

Spoiler: Height Measurement Custom Stage

Most characters have animation cycles that adjust their height. For those characters, low and high heights were recorded and an average was taken. If that average tied with another character, the character with the higher variance was considered to be "taller."

Some characters had significant features that required characterizing two heights. For example, when King Dedede crouches, his head dips far lower than his hand. His crouching height is characterized twice for that reason.

This method for assessing character height is riddled with error. Heights were visually assessed with a ruler taken to my screen, against a custom stage, while I had to pause to catch them in different parts of their standing/crouching animations. I also do not know where the hitbox for each character ends, so I tried to estimate the "top" of each character.

I encourage discussion on how to gather more accurate height data. This is all preliminary, and it needs revision.

RESULTS​

As a reminder, all heights are in "bricks," and some heights have higher significant digits due to it being the average of two representative animation cycle heights.

Spoiler: Character Height Ranking

Nice. I can use this to see who the second slash of Shulk's custom up B Advancing Air Slash (delayed) can hit while they are standing!

 

[Dr. Tuen]

Wouldn't it be a lot better to gather this data using a capture card? Then you can just have a fixed camera and measure whichever frame you want in pixels without pausing which seems much more accurate. Now I'm guessing you don't have a capture card, so that's obviously a problem. I could probably borrow my friend's capture card to record all stand and crouch animations in 1080p 60fps if that would be helpful.

Fixed camera perspectives are actually pretty difficult to work with. If you use the fixed camera feature in-game, it's far too zoomed out to get any useful data on relative height differences. If you use the game's zoom features and try to say... use maximum zoom to try and set a standard between characters it doesn't work out great either since larger characters push the camera out.

 

[kenniky]

Requesting height above ledge for characters? that was badly worded, sorry. What I meant is some characters, like Charizard, poke their head above the ledge. It might be useful to get that data as well.

 

[Dr. Tuen]

Requesting height above ledge for characters? that was badly worded, sorry. What I meant is some characters, like Charizard, poke their head above the ledge. It might be useful to get that data as well.

I believe you mean the height that character hitboxes reach over the edge when they grab it. A character dangling from a tether grab would have no hitbox near the edge, but Charizard (as you mentioned) has his head over it.

I'll look into it. I'm not exactly sure how to measure each character when it comes to ledge grabs... and if they also have movement cycles that effect their height I'll be pretty upset, ha ha. As a note to myself, this could be a good place to use giant smash. I also do these projects, alongside ZSS-specific projects, in my free time when I'm not working on my PhD... so we'll see! I'll add it to the list.

 

[kenniky]

I believe you mean the height that character hitboxes reach over the edge when they grab it. A character dangling from a tether grab would have no hitbox near the edge, but Charizard (as you mentioned) has his head over it.

I'll look into it. I'm not exactly sure how to measure each character when it comes to ledge grabs... and if they also have movement cycles that effect their height I'll be pretty upset, ha ha. As a note to myself, this could be a good place to use giant smash. I also do these projects, alongside ZSS-specific projects, in my free time when I'm not working on my PhD... so we'll see! I'll add it to the list.

Cool!

I can definitively say that the following characters' hitboxes reach the farthest above the ledge:

Yoshi (nose)
Bowser Jr.
DK
Meta Knight
Charizard
Ness
Villager
Olimar
Wii Fit Trainer WHEN HE/SHE BRINGS HIS/HER HEAD UP

as these can all be hit by the front hitbox of Shulk's utilt.

Wii Fit Trainer appears to be the only character whose movement cycle affects his/her height. You gotta fit in chin-ups somewhere in that moveset!

 

[#HBC | ZoZo]

Fwiw soem chars hitboxes change while crawling, maybe thats worth adding too?

 

[PUK]

Hilarious, and relevant for projectiles (ha). Though I think she can crouch under all of them...

Unfortunately no. Fully Aurasphere, charge shot etc... Have all a hitbox which hit slightly inside the stage to hit beneath the ledge (it a choice of game design) and because of this no crouch can help. Plus some fruit, gyro and laser (can be aimed), arrow and bomb all are affected by gravity so it can hit.
But most grab and Fsmash just whiff, and a lot of projectile too.

 

[Indigo Jeans]

Didn't see it mentioned but Zero Suit Samus is actually lower to the ground when crawling compared to when she's just crouching.

 

[evmaxy54]

I needed this in my life thank you very much gurl!

 

[FullMoon]

Greninja is the shortest one when standing? Wha.

 

[ThirdDay]

INTRODUCTION​

The purpose of this thread is to release and review preliminary character height data. These finding include character standing heights and character crouching heights. This is useful for determining who can be struck by falling aerials, rising aerials, and various grounded moves. While most important for the shortest characters in the game (Kirby, Pikachu, etc), some may find other application for this data.

While this will be discussed further, it is worth saying here: This is very rough data. It was based on observation. There are many sources of error.

Methodology​

A custom stage was created in order to have a wall to measure height by. That wall had bricks, and character height was estimated based on their position with respect to those bricks. Those estimations were used to create a preliminary height order. Any pair of characters that had matching or near matching height were compared directly and re-ordered if necessary.

Spoiler: Height Measurement Custom Stage

Most characters have animation cycles that adjust their height. For those characters, low and high heights were recorded and an average was taken. If that average tied with another character, the character with the higher variance was considered to be "taller."

Some characters had significant features that required characterizing two heights. For example, when King Dedede crouches, his head dips far lower than his hand. His crouching height is characterized twice for that reason.

This method for assessing character height is riddled with error. Heights were visually assessed with a ruler taken to my screen, against a custom stage, while I had to pause to catch them in different parts of their standing/crouching animations. I also do not know where the hitbox for each character ends, so I tried to estimate the "top" of each character.

I encourage discussion on how to gather more accurate height data. This is all preliminary, and it needs revision.

RESULTS​

As a reminder, all heights are in "bricks," and some heights have higher significant digits due to it being the average of two representative animation cycle heights.

Spoiler: Character Height Ranking

This includes a lot of good information. I'm glad you made this!

 

[E-Mann]

Very nice. I think overall, Rosalina is the largest female character, and Ganondorf is the biggest male character and overall.

 

[kenniky]

I saw this video:

This seems like a really consistent way to at least get a character's max height because hurtboxes can sometimes be deceiving.

 

[Locke 06]

When Mega crouches, his hands are at different heights. The hand behind him is tall enough for Rosa to standing grab it, but the hand in front of him is not. Not sure if that changes how you want to list Mega's stuff (or if head = front hand).

 

[Trieste SP]

Greninja is the shortest one when standing? Wha.

It surprised me as well but it makes sense considering his 'stand' looks a lot like a crouch.

 

[Shaya]

There's a way you can get the units of these standardized with in-game measurement.
I guess a little pot-luck, but finding the right example on mastercore of a move that's attached to bone 0 (i.e. centre of the model) displaced vertically enough that it's at their head height and then applying that scalar to the data you have.

Going by ZSS, I would guestimate that her height in game units is around 15-16.
(Her second hit of up tilt reaches to 23.5)

Her first hit of jab which I believe doesn't hit standing pikachu, is hitting as low as 9.1 units.

 

[E-Mann]

There's a way you can get the units of these standardized with in-game measurement.
I guess a little pot-luck, but finding the right example on mastercore of a move that's attached to bone 0 (i.e. centre of the model) displaced vertically enough that it's at their head height and then applying that scalar to the data you have.

Going by ZSS, I would guestimate that her height in game units is around 15-16.
(Her second hit of up tilt reaches to 23.5)

Her first hit of jab which I believe doesn't hit standing pikachu, is hitting as low as 9.1 units.

That's a decent way to do it, unless someone can actually rip the models in possibly a year from now.

 

[bc1910]

Greninja's crouch has hitbox issues.

It appears to be very short but he actually gets hit by a lot of crap that he should duck under.

It's most obvious with Falco's lasers. Go into training, pick Greninja and have CPU Falco, set camera to zoom, 1/4 speed and fire lasers at crouching Greninja. You'll see his head and body are clearly under the lasers but they literally hit the air above his head and he takes damage. I guess it kind of looks like they hit his ear but not really.

Greninja used to be able to crouch and crawl under Falco's lasers. I think his crouch was nerfed in 1.0.4 to be easier to hit. Various characters who clearly have taller crawls than Greninja, such as Diddy and Luigi, can crawl under Falco's lasers.

Don't know if this is just a Fox/Falco laser problem or affects his entire crouch. Probably the latter, but he can still duck under stuff like ZSS' jab which appears to hit the air above him. Will test.

His standing hitbox is fine and is indeed low, his stance makes him crouch like Wolf basically.

 

[Dr. Tuen]

Lasers may be an interesting way to update this work. Just an idea for me to look at later:

I could set up a stage where the left side contains a large number of platforms that are staggered and/or overlapped as close as possible to get small step sizes. Set one or two characters as fox and spam lasers. If the lasers hit the foe, move the fox(s) up a platform. If they miss, move down a platform. Once the minimum missed laser is found, pause the game as the laser is penetrating the wall that's set up on the right side of the stage. Measure from the bottom of the laser's animation, and use the same numbering system that's been used above.

Could be more accurate, will definitely take more time. And definitely must wait until after I defend my PhD.

Edit for my notes: use a large slanted platform to show for more precision.

 

[E-Mann]

Lasers may be an interesting way to update this work. Just an idea for me to look at later:

I could set up a stage where the left side contains a large number of platforms that are staggered and/or overlapped as close as possible to get small step sizes. Set one or two characters as fox and spam lasers. If the lasers hit the foe, move the fox(s) up a platform. If they miss, move down a platform. Once the minimum missed laser is found, pause the game as the laser is penetrating the wall that's set up on the right side of the stage. Measure from the bottom of the laser's animation, and use the same numbering system that's been used above.

Could be more accurate, will definitely take more time. And definitely must wait until after I defend my PhD.

Edit for my notes: use a large slanted platform to show for more precision.

I'll do it for ya if you want to.

 

[Dr. Tuen]

I'll do it for ya if you want to.

If you'd like to do that, go ahead. We can update here and I'll credit your work completely. If I find the time to do it first, I'll let you know if I've begun. Or, at least, I'll try to remember to.

Oh, and another methodological note:

The pattern on the wall I used is inconsistent. Some others have used the grid on the custom stage builder to place platforms at each grid mark, and at the midpoint of each grid mark. I'm sure even 1/4 platforms may fit too. This could provide an easier measurement system.

 

[E-Mann]

Lasers may be an interesting way to update this work. Just an idea for me to look at later:

I could set up a stage where the left side contains a large number of platforms that are staggered and/or overlapped as close as possible to get small step sizes. Set one or two characters as fox and spam lasers. If the lasers hit the foe, move the fox(s) up a platform. If they miss, move down a platform. Once the minimum missed laser is found, pause the game as the laser is penetrating the wall that's set up on the right side of the stage. Measure from the bottom of the laser's animation, and use the same numbering system that's been used above.

Could be more accurate, will definitely take more time. And definitely must wait until after I defend my PhD.

Edit for my notes: use a large slanted platform to show for more precision.

I... haven't even gotten around to doing that thing I promised yet, due to school.
I'll stay up overnight to finish the stage. And even then, it's got to be very precise, due to the characters not exactly standing still. Some do very dynamic poses, while others move up and down, if so slightly.

 

[Dr. Tuen]

I... haven't even gotten around to doing that thing I promised yet, due to school.
I'll stay up overnight to finish the stage. And even then, it's got to be very precise, due to the characters not exactly standing still. Some do very dynamic poses, while others move up and down, if so slightly.

I've tended to try and measure at the beginning of their stand cycle. Many characters stand in a pose, then move about a half second later. Others that fidget more can be stabilized by having them walk... but only if that doesn't effect their height too much.

 

[~ Gheb ~]

Some characters have noticeably lower height during their dashing/running animation than during their standing animation. There's probably no reliable way to measure it but I think for at least Lucina, Marth, Charizard, Pikachu, Little Mac, Sheik and Yoshi the difference is non-trivial. The same is true for certain dash attacks that allow tall characters like Rosalina to sneak in under moves that would otherwise hit her.

And since I've already started grasping at straws I'd also like to point out that crouching and crawling can be different things for some characters. The hurtbox of a crawling ZSS is lower than the hurtbox of a crouching ZSS. Again, probably not something that can be reliably measured.

 

[Dr. Tuen]

Update -

I tried to re-construct the stage from the video shown earlier in this thread and found that the lava in custom stages does not have the 100% kill property. This is unfortunate, as it means I can't reliably assess the time for a character to hit that surface (I don't possess the means to record at 60fps or do frame-by-frame playback).

I instead constructed the stage I described previously: a slope on the left side, thin platforms stacked on the right side as measurement. Fox fires lasers from the left and the character stands on the right. The place the laser passes through on the right indicates the measurement point. There is also a small bump added near the measurement platforms for characters to run, walk, or crawl into. This will allow me to take measurements for those behaviors too.

A few notes:

• Measurements taken this way are difficult, due to Fox's high run speed and my odd difficulty to get him to walk. I have found that him jumping on a slanted platform causes a short slide that can be useful for small adjustments.
• Characters have bouncy standing cycles. I may try to measure these and present the high/low/average values for them. This will extend the time of the project, but it will provide more accurate results.
• Similarly, there are crouch movement cycles too.
• Measurement of these cycles is a BENEFIT of the laser method over the lava-time method. The lava-time method, when done consistently, will reliably catch the same point in a character's crouch/stand movement cycle.
• On top of all of that, some characters have limbs that stick out in funny ways that effect height at different places on their bodies at different depths. For instance, Mega Man sticks one hand up when he crouches. If you are say... ZSS spacing bair, that hand won't come into play. If you are instead ZSS just trying to sweep Z air completely over him, that hand may determine whether or not he gets hit. These intricacies will need to be measured separately.
• Fractions of platforms are measured by measuring the entire platform height with a ruler on my screen (accurate to the millimeter), then measuring the location the bottom of the laser passes through, then dividing these two values. The reason for this type of measurement is due to the varying camera zoom. When paused, the camera will zoom in further when observing smaller characters. This means consistent measurement must be done on a relative scale.

I think that's it for now. I'll update again when I can.

 

[Masonomace]

Hey Tuen, sweet OP you have. I do have a question in regards to a tiny project I'm thinking about doing & it involves that question's answer.

Q: I'm testing one's jump height (let's say Shulk) & I want to find out which character would be best to test for a tapped footstool jump & a held footstool jump. Who would the best character example be? Someone like Greninja who's the shortest standing character who is still short crouching, or should it be someone like Kirby who is still short standing but is much shorter when crouching?

I learned that if I tap footstool jump on Zelda standing, & then tap foostool jumped on Zelda crouching, I indeed got a different footstool jump height when Zelda crouched. What kinda threw me off the first attempt I tried this was that Zelda's footstooled animation is the same whether she's standing or crouching, so I guess holding down on the joystick for Zelda crouching through the footstool does result with a shorter point to start. I'm thinking that perhaps it might be Kirby or Jigglypuff, but I don't know for sure.

 

[Dr. Tuen]

Hey Tuen, sweet OP you have. I do have a question in regards to a tiny project I'm thinking about doing & it involves that question's answer.

Q: I'm testing one's jump height (let's say Shulk) & I want to find out which character would be best to test for a tapped footstool jump & a held footstool jump. Who would the best character example be? Someone like Greninja who's the shortest standing character who is still short crouching, or should it be someone like Kirby who is still short standing but is much shorter when crouching?

I learned that if I tap footstool jump on Zelda standing, & then tap foostool jumped on Zelda crouching, I indeed got a different footstool jump height when Zelda crouched. What kinda threw me off the first attempt I tried this was that Zelda's footstooled animation is the same whether she's standing or crouching, so I guess holding down on the joystick for Zelda crouching through the footstool does result with a shorter point to start. I'm thinking that perhaps it might be Kirby or Jigglypuff, but I don't know for sure.

You are looking for Mega Man with an item in hand. When he holds an item, he stands perfectly still and lacks any animation frames.

This should provide a consistent basis for your footstool tests.

 

[Masonomace]

You are looking for Mega Man with an item in hand. When he holds an item, he stands perfectly still and lacks any animation frames.

This should provide a consistent basis for your footstool tests.

Excellent Tuen. Thanks to the max. Do ya think I should have Mega Man crouching or standing? He doesn't move when crouching with an item in hand so I may just do that.

 

[Deleted member]

If someone ripped all the models for either 3DS or Wii U, that could be a way to measure height.

 

[E-Mann]

If someone ripped all the models for either 3DS or Wii U, that could be a way to measure height.

Isn't that what I said a few weeks ago?

 

[MrNantendo]

I know this is an old thread but I wondered if this was going to get updated with the missing DLC characters?

 

[E-Mann]

I know this is an old thread but I wondered if this was going to get updated with the missing DLC characters?

The guy running this thread literally has so much work on his hands that it makes high school look like a walk in the park. Apparently, it's on hiatus until he decides to actually update the damn thing.

I will tell this, however, that Corrin is as tall as ZSS, and Bayonetta is the second tallest female in the game, behind Rosalina.