Ness's Back Throw and KO Percents - The Thread

[Noa.]

~This thread is a collection of data that details all the percents that bthrow KOs at~​

Introduction:

The table below shows at what percent each character dies from bthrow on Final Destination. The lowest percent was taken by using bthrow all the way on the edge, as close to the blast zone as possible while still being on Final Destination. The middle percent shows what percent bthrow KOs at the center of the stage. The highest KO percent is taken by using bthrow as far away from the blast zone as possible, on the opposite ledge on Final Destination. The fourth row of this table has a mean average of these three KOs, so you can more easily compare characters across the board.

The first column has each character ranked by their weight. Clicking on the column at the top will arrange the characters in order from lightest to heaviest, or heaviest to lightest. The second column has all of their names, and you can arrange them all alphabetically. The following columns have all of the different KO percents. Clicking on any of these columns will arrange the characters in order of KO percent for that specific column.

The testing for this table was done in training mode. Ness was used by Player 1 and the opposing character was controlled using another controller. There is no rage factored into these KO percents. These KO percents were calculated on Final Destination on the Wii U. DI is being used by the character being killed. For the closest KO and middle KO percents, the control stick was held in the opposite direction of bthrow. So if Ness is facing to the right and grabs the opponent, the control stick would be held in this direction: . The correct way to DI Ness's bthrow at these percents is to hold directly opposite to the direction that the character is being thrown. On the farthest KO however, this is not the most optimal way to DI. If the control stick is just held opposite to bthrow, the character will die off the top. The correct way to hold the control stick is downward, and diagonally in the opposite direction of bthrow. If Ness is facing to the right and grabs his opponent to bthrow him across the whole stage, the control stick should be held in this direction: . All of the characters were tested using this DI. The control stick was held in that direction before being thrown, and until the character was already visible on screen and outside of the bubble camera near the blast zone.

This DI did not use the c-stick, nor was there any attempt to cancel momentum by jumping, air dodging, or fastfalling an aerial. It is possible that doing any of these actions in tandem with the control stick DI would prolong survival. For this test, only DI with the control stick was used. The difference in KO percent that this DIing brings can range anywhere from 10% to 20%. I'm confident in the data of this table up to + or - 1%.

Here is the table:

Weight|Character|Closest|Middle|Farthest|Mean
0.|Jiggs|79|108|131|106
1.|Mewtwo|81|112|137|110
2.|GW|82|115|137|111.33333
3.|Kirby|82|115|138|113
4.|Rosalina|86|118|142|115.333
5.|Fox|84|117|139|113.3333
5.|Pikachu|84|117|141|114
5.|Olimar|85|118|142|115
8.|MetaKnight|84|118|141|114.3333
9.|ZSS|86|120|144|116.6666
10.|Falco|86|120|142|116
10.|Little Mac|86|120|143|116.33333
12.|Sheik|88|122|146|118.6666
12.|Zelda|89|123|148|120
14.|Peach|92|125|150|122.3333
15.|Smash Shulk|85|119|143|115.666
16.|Marth/Lucina|92|125|151|122.666
17.|Duck Hunt|92|127|152|123.666
17.|Palutena|93|127|152|124.333
19.|Diddy Kong|91|127|153|123.6666
19.|Toon Link|93|128|154|125
21.|Ness|93|129|155|125.66666
21.|Greninja|92|127|152|123.6666
23.|Pacman|94|130|155|126.333
23.|Sonic|93|129|155|125.666
23.|Robin|94|130|156|126.666
26.|Pit/Dark Pit|94|130|157|127
26.|Wii Fit|94|131|157|127.333
28.|Luigi|94|130|156|126.666
28.|Villager|94|131|157|127.333
30.|Mario/Doc|95|131|158|128
31.|Lucario|96|132|159|129
32.|Default Mii|95|132|159|128.666
33.|Megaman|98|136|161|131.666
33.|Shulk|97|134|161|130.666
35.|Yoshi|97|135|163|131.666
35.|Link|98|135|161|131.333
35.|Falcon|99|135|162|132
38.|ROB|97|135|162|131.333
39.|Wario|101|138|166|135
39.|Ike|100|137|165|134
41.|Samus|102|141|167|136.666
42.|Ganondorf|102|140|168|136.666
43.|Bowser Jr.|99|137|165|133.666
43.|Charizard|104|143|172|139.666
45.|Dedede|104|144|175|141
46.|Donkey Kong|105|146|176|142.333
47.|Bowser|109|150|182|147
48.|Shield Shulk|122|186|221|176.333

The tables below show the mean for the KO percents at each distance across all the characters, as well as the standard deviation for each distance. The first table includes all the data, while the second table does not include shield Shulk nor smash Shulk in its calculations.

Distance|Mean|Standard Deviation
Closest|93.5102|8.08322
Middle|129.69388|12.3764
Farthest|155.6939|14.7411

Without Shield Shulk or Smash Shulk

Distance|Mean|Standard Deviation
Closest|93.08511|6.98383
Middle|128.7234|9.35229
Farthest|154.57447|11.36881

Analysis:

The percentage at which Bthrow KOs is determined by two attributes of the character; their weight and fall speed. Weight has far more of an impact on when a character dies. That's why when you arrange the table in order of weight, it's very similar to the order in which the characters die to bthrow. However, when characters are the same weight, or have a similar weight, fall speed can actually be strong enough to affect when a character dies. Characters with a faster fall speed die at a slightly lower percent, while floatier characters can survive a bit longer. The effect that fall speed has on bthrow KO percents is small, but significant enough to affect the KO by a handful of percents.

The standard deviations also reveal some interesting information about the KO percents of bthrow. Here is the definition of standard deviation for those that are not familiar with the term.

standard deviation: a quantity calculated to indicate the extent of deviation for a group as a whole.

Essentially, the standard deviation measures how much variation there is in a group of data.

Looking at our information here, we can see that the standard deviation for the KO percents closest to the blast zone is smaller than the standard deviation for the KO percents farthest from the blast zone. The farther that a character has to travel with bthrow, the larger an effect their weight has. What this means to you as a player is that when you grab your opponent closer to the blast zone, you can place more significance on their percent rather than their weight when you're determining whether or not bthrow will kill. Conversely, when you have to bthrow your opponent across the stage, it is a lot more difficult to kill heavier opponents. The difference in the variation between the closest and farthest distance is not huge, but quite noticeable. Make sure to take it into account.

Next, we can separate our list of characters into groups of different kill classes. This was done by taking the mean KO percent of all of the characters, and creating a gap between any characters that had a difference greater than 2%.

Always Dies to Bthrow:
Jiggs

Almost as Light as JIggs:
GW

The Lightweights:
Kirby
Fox
Pikachu
MetaKnight
Olimar
Rosalina
Smash Shulk
Falco
Little Mac
ZSS

Only Separated by their Fall Speed:
Sheik
Zelda

A Huge Collection of Middleweights:
Peach
Marth/Lucina
Duck Hunt
Diddy Kong
Greninja
Palutena
Toon Link
Sonic
Ness
Pacman
Robin
Luigi
Pit/Dark Pit
Wii Fit
Villager
Mario/Doc
Default Mii
Lucario

Heavy enough to make a difference:
Shulk
ROB
Link
Yoshi
Megaman
Falcon

The Light Heavyweights:
Bowser Jr.
Ike
Wario

Samus and Ganon (No Name for this One):
Samus
Ganondorf

The Fatties:
Charizard
Dedede
Donkey Kong

The Head Fatty in Charge:
Bowser

Just Wait out the Transformation ;_; :
Shield Shulk

All the characters in each group die at roughly the same percents due to their combination of weight and fall speed. The only exception would be the huge collection of middleweights. Their mean KO percents range from 122% to 129%, which is a range of 7%. The scaling in this group is incremental, and there was no breaking point to separate this group into two. Also, don't try to kill shield Shulk. Seriously.

Rage:

In addition to our KO percents with no rage, we also ran some tests with rage to see its effect on our Bthrow KO percent. Our testing with rage is not as in depth, because it is difficult to test the effects of rage and very time consuming. Here is our data:

Ness with 150% Rage

Character|Closest|Center|Farthest
Jiggs|62|86|105
Greninja|73|103|123
Bowser|90|121|149

Ness with No Rage

Character|Closest|Center|Farthest
Jiggs|79|108|131
Greninja|92|127|152
Bowser|109|150|182

Difference in Percent with 150% Rage and No Rage

Character|Closest|Center|Farthest
Jiggs|17|22|26
Greninja|19|24|29
Bowser|19|29|33

Looking at these tables, it's easy to see what kind of effect distance has on how impactful rage can be. Whenever characters have to travel farther for bthrow, they die significantly earlier when Ness has rage. Rage caps out at 150%, so this is the biggest effect that rage can have on kill percents.

The table below shows what percents different characters die at with different amounts of rage. All of the data was done on the wii u on final destination, with the same DI described above. All of these percents are from the center of Final Destination. There is a small difference between the data on Jiggs and Bowser from the table above and the table below. This is due to small amounts of human error. The amount of error in this data is at a much higher margin than the data taken earlier for the KO percent of each character.

Character|No Rage|90% Rage|110% Rage|130% Rage|150% Rage
Jiggs|108|96|93|90|88
Rosalina|118|106|100|98|95
Sheik|122|107|104|102|98
Diddy|127|112|108|104|102
Mario|131|117|115|110|107
Yoshi|135|119|116|111|110
Bowser|150|137|130|126|123

Difference Between KO Percent with No Rage and Rage

Character|No Rage|90% Rage|110% Rage|130% Rage|150% Rage
Jiggs|0|12|15|18|20
Rosalina|0|12|18|20|23
Sheik|0|15|18|20|24
Diddy|0|15|19|23|25
Mario|0|14|16|21|24
Yoshi|0|16|19|24|25
Bowser|0|13|20|24|27

With the exception of a couple of outliers due to human error, you can see that rage has a greater effect on heavier characters. The difference in KO percents is bigger on the heavier characters, especially when you get to higher amounts of rage.

That marks the end of the data on Bthrow that's been collected so far! Hopefully you have a better idea of how character weight, character fall speed, stage position, DI, staleness and rage all affect the KO percents of bthrow. If you have any other ideas of what should be tested, suggest them. There are also probably lots of both mathematical errors and human errors in this, so if you see anything suspicious validate it. We triple checked a lot of the data on here, but the power of human error is great.

All of the physical testing, analysis, and construction of testing processes were done by Luco and Draconoa44.

 

[Noa.]

Luco and I collaborated on this thread together. It took a while to make and was a lot of work. I hope everyone appreciates it. If you have any suggestions or questions, please feel free to say something!

 

[Teran]

This is awesome stuff and all character boards should do things like this.

Also puff puff Jigglypuff, Luco will get it

 

[PKBeam]

WFT's deep breathing makes her heavier.
also, nice job

 

[Noa.]

Maybe I'll add her in. One day. If she ever escapes low tier.

 

[PKBeam]

if you look around on WTF boards there is a certain person who has an unusual opinion on her placing.
the same person who said she wins against Ness.

 

[Luco]

Credits also go to @ Teran for helping test some Jigglypuff-related rage kill percents and the effects of staling on Bthrow which turned out to have little effect on these results because it's hard to stale Bthrow to a point where these percentages no longer apply :3

 

[Snerp]

Great stuff guys. I'm just waiting for that one person to chime in, crying about bthrow.

 

[Pazx]

This move is filthy. I'm saving this thread.

By the way, I applaud your thoroughness, your writeup and your method. Well done.

 

[PKBeam]

Bthrow is the cleanest kill Ness can get. No whiffing it.

 

[Earthbound360]

This is better than oxygen. The content is extremely useful, the pictures are a nice touch, it's well researched, and organized well. You've gotta tell me how you did that "rearrange the chart" thing.

You two rock.

 

[The 0ne]

Fantastic work guys! This is one of those things that's like I didn't know I wanted it before I knew about it. Extremely detailed and useful; definitely going to help me out when b-throwing my opponents. And damn, I didn't know Samus was THAT heavy.

Maybe add in Shield and Smash Kirby? You know for teams, I guess.

 

[CURRY]

Dayum, killing Shield Shulk at 122% at the ledge?
That's actually SUPER early, lol. Take that opportunity if you can, if it's early in the transformation or something.
I play Shulk sometimes... I'm having trouble choosing a Smash 4 main, lol. But if a Ness killed me in my Shield mode at that low of a percent, it would be SO unexpected, since I expect to be pretty much invincible until the 200%s, unless my opponent is going really aggressive for the gimps.
Then again, the gimps. Shulk has a meh recovery. That might be a LITTLE more efficient than fishing for a bthrow.

 

[Earthbound360]

Oh, something that you guys may want to consider adding is pummel additions. The number of pummels you can get in are correlated by the damage that the opponent has when you grab them. Ness' pummel does 1% per hit, so it could effectively push the damage just above that KO threshold.

You could do testing to see just about how many pummels you can get in at certain percents, ultimately lowering what percents you could KO a character at any given time (from the point of the grab). This may be hard to test however, since mashing is something that's difficult to get consistent testing on, and people mash to break grabs sooner.

From my experiences though, my general rule is something like this:

0%-20% = No pummeling
21%-30% = 1 pummel
31%-99% = 2 pummels
100%-140% = 3 pummels
141%+ = 4 pummels

From my experience, the people who have mashed the fastest out of my grabs will always take the pummels AND the throw. Sure, you can add a pummel here and there if you're willing to risk it, but for guaranteed damage+throw, this is what I do.

 

[Funen1]

This may be my new favorite tech thread. Really fantastic works, you guys.

I do have a question about the description of the DI you said you used. My understanding of DI in the current iteration of Smash 4 is that it's the same (or almost the same) "vectoring" as in the original 3DS version, but that you cannot vector downward anymore (basically, the upper half of the region between left and right still works). However, I've been having trouble finding reliable answers regarding this topic, outside of Ninjalink's videos (I think they were his anyway, correct me if I'm wrong). If I'm missing something here, I'd greatly appreciate it if someone would clarify it. Would holding down for the "far side of FD" B-throws perhaps have something to do with activating fast-falls, and that's what saves you, or does it still count as part of the DI calculation?

 

[PKBeam]

traditional DI is still in Smash 4, just almost nonexistent for high angles e.g. Rest, where it impacts about 5% at most.
Holding down while shooting across FD means that your DI angle is lower so you don't pass the top blastzone too early.
I'm not sure about fastfalls. Might be both.

also, because of the way Smash 4 recognises inputs, spinning any joystick is the fastest way to break out of grabs' shield break stun, sing etc. because the game checks at each frame for an input, and multiple inputs during one frame count as one input.

 

[Nd_KakaKhakis]

If people explored other moves in this game with this thoroughness this game would progress at lightspeed.

Great thread.

 

[Funen1]

traditional DI is still in Smash 4, just almost nonexistent for high angles e.g. Rest, where it impacts about 5% at most.
Holding down while shooting across FD means that your DI angle is lower so you don't pass the top blastzone too early.
I'm not sure about fastfalls. Might be both.

also, because of the way Smash 4 recognises inputs, spinning any joystick is the fastest way to break out of grabs' shield break stun, sing etc. because the game checks at each frame for an input, and multiple inputs during one frame count as one input.

Ah, I gotcha. Makes much more sense now. Thanks for the help.

 

[Sir_Zedd]

This thread is spectacular, looks like a lot of work has gone into it and its paid off.

Ness's everywhere will be appreciating this information, thankyou both.

 

[Noa.]

This is better than oxygen. The content is extremely useful, the pictures are a nice touch, it's well researched, and organized well. You've gotta tell me how you did that "rearrange the chart" thing.

You two rock.

Whenever you make a table on smashboards, if you put table head in the first line, it'll create labels for each column. Then you can just click on the column title to rearrange the contents.

Oh, something that you guys may want to consider adding is pummel additions. The number of pummels you can get in are correlated by the damage that the opponent has when you grab them. Ness' pummel does 1% per hit, so it could effectively push the damage just above that KO threshold.

You could do testing to see just about how many pummels you can get in at certain percents, ultimately lowering what percents you could KO a character at any given time (from the point of the grab). This may be hard to test however, since mashing is something that's difficult to get consistent testing on, and people mash to break grabs sooner.

From my experiences though, my general rule is something like this:

0%-20% = No pummeling
21%-30% = 1 pummel
31%-99% = 2 pummels
100%-140% = 3 pummels
141%+ = 4 pummels

From my experience, the people who have mashed the fastest out of my grabs will always take the pummels AND the throw. Sure, you can add a pummel here and there if you're willing to risk it, but for guaranteed damage+throw, this is what I do.

I think testing how many pummels you can get against someone who is mashing their controller correctly would be useful. I'll see if I can do that some time.

 

[PKBeam]

so FYI the best way to "mash" is to just spin a joystick around. it counts as an input on every frame.
there's no point in pressing buttons or doing anything else if you just spin a joystick, since the game only accepts one input per frame, you just gonna break your controller and have to get another one.

 

[Noa.]

But how can I strike fear into opponent's heart with my awesome tech skill? They have to hear me.

 

[CURRY]

I was about to comment on how there was almost no difference between shield break stun time when I spun both the analog and the c-stick around at the same time when I was testing stuff yesterday.
Then I remembered how the c-stick actually overrides the analog controls.
Ohhhh, how stupid I am.

 

[PKBeam]

But how can I strike fear into opponent's heart with my awesome tech skill? They have to hear me.

How can you strike fear into your opponent's heart when your controller breaks midmatch and you see them pushing you to the ledge and jabbing you offstage?

 

[Noa.]

How can you strike fear into your opponent's heart when your controller breaks midmatch and you see them pushing you to the ledge and jabbing you offstage?

I don't know about you but my controller is made from the fiery depths of Mt. Doom, and is too tough to break.

 

[PKBeam]

I don't know about you but my controller is made from the fiery depths of Mt. Doom, and is too tough to break.

Makes it all the more sad when it DOES break,,,

 

[Jibbablabba]

>playing For Glory
>A wild Ness main appears
>D3 selected
>backthrow into the promised land
>i learned something that day...that Bthrow is scary AF

 

[PixelPerfect]

This is really useful information. Nice job! We need some other threads with a detailed look at certain moves like this.

 

[Luco]

>backthrow into the promised land

Oh hey, I'm there right now! I'll look out for a depressed fat penguin next time I'm out walking in Jerusalem!

But in all seriousness @ Noa. , if you're looking for a few more things to test out (as we were discussing), my suggestion would be to do a thorough test of the legitimacy of our combos (Dthrow--> Fair, 0%-60% combos abusing invincibility and PK Fire off the ledge ---> DJ Fair ---> FF Uair ---> SH Nair ---> grab and Dthrow ---> double/triple Fair combo ((... this one is hard to pull off but really satisfying when it works!)) combos at different percents and how vectoring/DI affects the combo, etc). Also if ya need any help with anything, shoot me a PM and I'll do my best to help from over here =P

 

[Noa.]

Oh hey, I'm there right now! I'll look out for a depressed fat penguin next time I'm out walking in Jerusalem!

But in all seriousness @ Noa. , if you're looking for a few more things to test out (as we were discussing), my suggestion would be to do a thorough test of the legitimacy of our combos (Dthrow--> Fair, 0%-60% combos abusing invincibility and PK Fire off the ledge ---> DJ Fair ---> FF Uair ---> SH Nair ---> grab and Dthrow ---> double/triple Fair combo ((... this one is hard to pull off but really satisfying when it works!)) combos at different percents and how vectoring/DI affects the combo, etc). Also if ya need any help with anything, shoot me a PM and I'll do my best to help from over here =P

Honestly I think my next project is gonna be PK fire. I wanna test how DI, damage, distance the lightning bolt flies, and where it makes contact affects the options of pk fire. And testing how often opponents can shield, spotdodge, roll, jump, jab, and any special moves that can be used while in pk fire.

 

[PKBeam]

just so you know the first hit of PKF is SDI 0x & the rest are SDI 0.8x

 

[The PK Dragon]

I don't know about you but my controller is made from the fiery depths of Mt. Doom, and is too tough to break.

that Lord Of The Rings reference made my day xD

 

[NorCal_DSS]

This is extremely useful. Sharing it with the NorCal smash page. Thanks for collecting this data.

 

[Luco]

Hey @ Noa. when you get the time, would you be up to adding in Mewtwo's results to the table? The reason I suggest doing it now and not when Lucas is released is partially to keep this thread relevant/easily found (even with the thread directory) and also cause it'll be handy to have the data as early as possible. I would do it myself but I only have 3DS which affects the kill percentages ( ). If you want to do it and have a bit of spare time, it might be handy to include Mewtwo in the rage-affected kill %s table. I'm pretty sure he has a weight of 72, the second lightest vanilla char in the game just above puff, but I can check. :3

Um, but don't feel the pressure to do it right now or anything, there's no real rush. =D

EDIT: Can confirm that Mewtwo's weight is 72, making him the second lightest in the game after Jiggs.

 

[Noa.]

I can't do it now cause I don't own mewtwo. I forgot to register my copies of smash. I can do it when mewtwo comes out for everyone.

 

[Meccs]

Mewtwo data:
Close: 81
Middle: 112
Far: 137

 

[Noa.]

Mewtwo data:
Close: 81
Middle: 112
Far: 137

Thank you for doing this. This data looks accurate. I added it to the chart. I just want to ask, did you do this with proper DI?

 

[Meccs]

held the opposite direction on close and middle, held opposite and down on the farthest.

 

[The 0ne]

Hey @ Luco how exactly do you determine a character's weight?

 

[Luco]

Hmm, I'm not too sure myself, I mean you can always determine who they're lighter than/heavier than by doing a multitude of KO percents testing, but for the actual numbers I suspect people look at game data to figure it all out, because the game would include numbers to represent weight somewhere in that coding. I know they've mined exact data recently but I'm not sure whether the numbers we used at the beginning were the actual numbers the game used or not ahah. :3

Oh, and finally G&W has someone to join him on the 'almost as light as jiggs' category!