Q&A Mechanics & Techniques Discussion

[TheReflexWonder]

Glancing at the numbers, I seem to have found the (rather obvious, in retrospect) single most significant factor in determining how much Rage will affect a move...Base Knockback.

Looking over some past Reddit stuff, as well as the frame data repository--

http://www.reddit.com/r/smashbros/comments/2r9mc3/smash_4_an_indepth_look_at_the_rage_and_how_it/
http://www.reddit.com/r/smashbros/comments/2rc953/smash_4_further_rage_testing_falcons_side_b_diddy/
http://pastebin.com/PzTjdtx8

Among the highest:

Captain Falcon Raptor Boost (Side B)
No rage - Kills Ike at 127%
50% rage - Kills Ike at 119% [0.16 DPR, Reduces damage needed by 6.3%]
100% rage - Kills Ike at 99% [0.28 DPR, Reduces damage needed by 22.0%]
150% rage - Kills Ike at 83% [0.293 DPR, Reduces damage needed by 34.6%]

Raptor Boost (ground)
Frame 4- 8: 9%(+4) 100b/75g (KO@ 150%) 85° 2.0-Hitlag Fire
Max Damage: 9%

Pit Upperdash Arm (Side B)
No rage - Kills Ike at 140%
50% rage - Kills Ike at 133% [0.14 DPR, Reduces damage needed by 5%]
100% rage - Kills Ike at 112% [0.28 DPR, Reduces damage needed by 20%]
150% rage - Kills Ike at 93% [.313 DPR, Reduces damage needed by 33.6%]

unknown substate 0x254
Frame 2- 4: 11%(+10) 100b/60g (KO@ 160%) 80° 2.0-Hitlag
Max Damage: 11%

Diddy downthrow
No rage - Ike reaches platform at 94%
50% rage - Ike reaches platform at 86% [0.16 DPR, Reduces damage needed by 8.5%]
100% rage - Ike reaches platform at 56% [0.28 DPR, Reduces damage needed by 29.8%]
150% rage - Ike reaches platform at 34% [0.4 DPR, Reduces damage needed by 63.8%]

D-throw
Frame 1- 2: 7% 85b/40g (KO@ 435%) 105°
Max Damage: 7%

And on the other end of the spectrum--

Toon Link Upair, Battlefield center platform
No rage - Kills Ike at 114%
50% rage - Kills Ike at 110% [0.08 DPR, Reduces damage needed to kill by 3.5%]
100% rage - Kills Ike at 100% [0.14 DPR, Reduces damage needed to kill by 12.3%]
150% rage - Kills Ike at 92% [0.147 DPR, Reduces damage needed to kill by 19.3%]

Uair
Frame 11-13: 14% 25b/95g (KO@ 144%) 80° 1.1-Hitlag Pierce
Frame 14-40: 11% 25b/95g (KO@ 144%) 80° 1.1-Hitlag Pierce
Max Damage: 14%

Ness bthrow, far edge
No rage - Kills Ike at 145%
50% rage - Kills Ike at 140% [0.1 DPR, Reduces damage needed by 3.4%]
100% rage - Kills Ike at 130% [0.15 DPR, Reduces damage needed by 10.3%]
150% rage - Kills Ike at 122% [0.153 DPR, Reduces damage needed by 15.9%]

B-throw
Frame 1- 2: 11% 15b/130g (KO@ 133%) 135°
Max Damage: 11%

These moves run the gamut in terms of potential confounds that don't seem to apply--It appears to be consistent between horizontal and vertical attack options, throws and direct attacks, high damage and low damage.

I wonder if we could piece together a formula based on that, or at least a theory as to what Rage really does. My guess is that it's a Base Knockback multiplier.

 

[DestinNotDustin]

I haven't kept up with Smash 4 news lately, so hearing about a Rage (or comeback) mechanic is a pleasant surprise.
Keep digging! Can't wait to see what else you find!

 

[popsofctown]

Could the data possibly instead mean that rage improves knockback growth by a fixed value? That would be a six of one half a dozen of the other interpretation I think. Either rage doesn't help low base high growth moves because it improves a % of the base kb or because it improves a fixed amount of growth (which in terms of % total knockback will be related to fixed divided by natural growth, so you're looking at either base*rage or rage*(1/growth) )


This post might be unintelligible, sorry if it is.

 

[TheReflexWonder]

Could the data possibly instead mean that rage improves knockback growth by a fixed value? That would be a six of one half a dozen of the other interpretation I think. Either rage doesn't help low base high growth moves because it improves a % of the base kb or because it improves a fixed amount of growth (which in terms of % total knockback will be related to fixed divided by natural growth, so you're looking at either base*rage or growth*(1/rage) )


This post might be unintelligible, sorry if it is.

It's likely not a fixed value directly applied to KBG, as moves with excessively low KBG, such as Villager and Ness F-Throws, won't KO at any reasonable percent, despite their abnormally high BKB. As always, BKB and KBG come together to figure out how far they go...I believe KBG is tied much more strongly to damage dealt on that attack, though, while BKB isn't weighed down by low damage as much.

 

[DanGR]

Glancing at the numbers, I seem to have found the (rather obvious, in retrospect) single most significant factor in determining how much Rage will affect a move...Base Knockback.

No wonder Luma up-air is so strong at rage percents.

110b/60g

 

[LimitCrown]

What if it's just a knockback multiplier that affects the entire knockback value? It would directly affect the base knockback value and the component of knockback that scales with damage simultaneously.

 

[TheReflexWonder]

What if it's just a knockback multiplier that affects the entire knockback value? It would directly affect the base knockback value and the component of knockback that scales with damage simultaneously.

If that were the case, we'd likely see a more pronounced effect at higher damage values for high KBG/low BKB moves, as without Rage, KBG affects knockback much more heavily at KO percents than BKB (with the exception of moves with almost zero KBG).

In the grand scheme of things, it still all comes together; Ness F-Throw isn't KOing at reasonable percents regardless of Rage despite its incredible BKB due to its incredibly low KBG. The fact remains that BKB appears to be (and by far) the most relevant value.

 

[LimitCrown]

Is rage a continuous function? More specifically, is there a gradual increase of the amount of knockback dealt as the damage sustained by the user increases?

 

[TheReflexWonder]

Is there a gradual increase of the amount of knockback dealt as the damage sustained by the user increases?

Yes, though I do not know the differences from percent to percent (is the difference between, say 0% Rage and 50% Rage always smaller than the difference between 50% and 100%?). It seems to cap at around 150%.

 

[LimitCrown]

Yes, though I do not know the differences from percent to percent (is the difference between, say 0% Rage and 50% Rage always smaller than the difference between 50% and 100%?). It seems to cap at around 150%.

I'm currently calculating the the amount of knockback that is dealt by some of the attacks at those damage percentages. It seems that the knockback difference between 0% and 50% is much smaller than the difference between 50% and 100%.

EDIT: I've calculated the amount of knockback that was dealt by Captain Falcon's Raptor Boost, Ness's backward throw, and Toon Link's up air attack. If I assume that the entire knockback value is altered by the rage multiplier, then the results for the possible multiplier values seem to be relatively consistent. Any differences between the ratios for each of the attacks may be due to an error in how the data was collected. However, if I assume that only the base knockback value is altered by the rage multiplier and try to calculate the multiplier's values, then the results don't seem to be very consistent. Therefore, it's more likely that the multiplier affects the final knockback value. I estimate the value of the rage multiplier at 0%, 50%, 100%, and 150% to be 1, 1.025, 1.1, and 1.15, respectively.

 

[Big O]

Rage is a total knockback multiplier. The reason rage "scales" best with base kb is because of simple math. Let's say it takes 200 kb to KO someone. Let's assume for simplicity that kb = base kb + kbg/100 * damage. Move A has 100 base kb and 100 kbg, so it KO's at 100% (100 + 100/100 * 100 = 200). Move B has 150 base kb and 50 kbg, so it also KO's at 100% (150 + 50/100 * 100).

However when you add a multiplicative factor to it, they reach the 200 kb threshold at different %'s. At max rage the multiplier seems to be pretty close to 1.15, so let's just use that for the math. 200 = 1.15(100 + 100/100 * A). 200 = 1.15(150 + 50/100 * B). The damage needed for A to KO is 74%. The damage needed for B to KO is 48%. A 26% difference despite both KO'ing at 100% without rage.

If you play around with all the numbers you will notice a clear pattern for how they affect the difference in KO%'s. Basically as kbg gets smaller and as bkb gets bigger, the % difference for KO's due to rage will increase. Also the scariest character in the game with rage is Luma. They probably switched the numbers for bkb and kbg by accident or something which lead to that monstrosity lol. That's my guess anyway.

As for how rage scales with %, I think some of the people that have tested it came to the conclusion that it doesn't start until like 30 something % and ends at 150%. I'd wager it scales linearly starting from that point until 150.

 

[LimitCrown]

I'm not sure whether there is a single linear function that is used for damage values between 30% and 150% or whether there are multiple linear functions that are used within that damage range. In the former case, at 50% the multiplier's value would be exactly 1.025. At 100%, however, the value would be 1.0875.

If the formula is only a piecewise function composed of only three separate linear equations, then this is possibly the formula, assuming that the value of the multiplier begins to change when the amount of damage that the character sustains is over 30%:

r = 100 if d ≤ 30
d/8 + 385/4 if 30 ≤ d ≤ 150
115 if d ≥ 150

The variable r is basically the value of the rage multiplier multiplied by 100.

 

[Mikkelmann]

Hey I found a falcon wavedash like movement. The quality is not very good but I'll make sure to get a better video up sorry my capture card is bad.

Video 1

Video 2

Let me know if this is found or it's already known.

How To Do​

1. You need the custom move "Wind-Up Raptor Boost"

2. Go to a slanted platform or a stage area

3. Side+B away from it and you should do a slide

NOTE: if you do it like a wave land you will slide farther. And it can be done on the ground but it's hard and still testing.

PLEASE DON'T POST ON REDDIT. And I may edit the thread for corrections. And sorry if I mess up or it takes a while for me to show it again like I said an other video is coming soon.

NOTE: You do NOT need to jump to do this. On some stages you do like Kongo Jungle but on stages like Lylat you can do it run back up and do it again. And you can act out of it.

 

[Azazel]

Hey I found a falcon wavedash like movement. The quality is not very good but I'll make sure to get a better video up sorry my capture card is bad.

Video 1

Video 2

Let me know if this is found or it's already known.

How To Do​

1. You need the custom move "Wind-Up Raptor Boost"

2. Go to a slanted platform or a stage area

3. Side+B away from it and you should do a slide

NOTE: if you do it like a wave land you will slide farther. And it can be done on the ground but it's hard and still testing.

PLEASE DON'T POST ON REDDIT. And I may edit the thread for corrections. And sorry if I mess up or it takes a while for me to show it again like I said an other video is coming soon.

NOTE: You do NOT need to jump to do this. On some stages you do like Kongo Jungle but on stages like Lylat you can do it run back up and do it again. And you can act out of it.

It has already been found.

Spoiler

Certain specials involving movement can cause rather significant slides when cancelled by slopes. @DatRedYoshi posted about this with (Dark) Pit's sideB yesterday; I did some brief experimentation and found Captain Falcon's sideB 3 and, in a less similar vein, Robin's upB 3 can be used for this as well.

With the Pits/Falcon, you jump and immediately sideB away from the slope.

Not sure if there's a whole lot of use for this (can't act until near the end of the slide), but at the very least it's a neat little movement gimmick. I'd imagine more characters can do something like this too.

 

[Mikkelmann]

It has already been found.

Oh ok can you send me a link so I can get more info on it?

 

[Azazel]

Oh ok can you send me a link so I can get more info on it?

It was in this thread on page 18

 

[RavensArk]

I'm pretty sure this is where it belongs, also I don't see a lot of people talking about this so I thought I'd post a thread. Basically while messing around in training mode I found out I could delay my dash grab. Dashing stopping and waiting for less then a second allows you to delay your dash grab, essentially making it a good mind game tool. I thought this was cool, I haven't seen many people use this so I thought I'd share.

 

[Gunla]

I'm pretty sure this is where it belongs, also I don't see a lot of people talking about this so I thought I'd post a thread. Basically while messing around in training mode I found out I could delay my dash grab. Dashing stopping and waiting for less then a second allows you to delay your dash grab, essentially making it a good mind game tool. I thought this was cool, I haven't seen many people use this so I thought I'd share.

Moved to this thread. Just notifying you.

 

[LimitCrown]

When would it be good to crouch cancel attacks in Smash 4? It certainly isn't as strong as it was in Melee, but it is stronger than it was in Brawl.

 

[Quillion]

So has anyone already definitively figured out how hitstun works in this game? It's likely tied to knockback just like in other games, but what are the details?

 

[Shaya]

Different hitbox types may have different types of hitstun (ZSS paralyzer stuff), but otherwise they all scale with knockback (and the only distinction we're aware of is pre-tumble and post-tumble percent [pre-tumble was maximum 30 frames of hit stun in Brawl]). If you ever pay attention to the dust clouds that follow you after being hit, they dissipate just as hit stun ends.

 

[Quillion]

Different hitbox types may have different types of hitstun (ZSS paralyzer stuff), but otherwise they all scale with knockback (and the only distinction we're aware of is pre-tumble and post-tumble percent [pre-tumble was maximum 30 frames of hit stun in Brawl]). If you ever pay attention to the dust clouds that follow you after being hit, they dissipate just as hit stun ends.

Oh, God, I thought so.

I really hope they don't take this "hitstun multiplier" approach in future Smash games with Sakurai's moving on. It's a lazy and inelegant piece of programming that doesn't lead to any good balancing (and even Project M will have to deal with that for all eternity since they can't even change it).

 

[Shaya]

I don't know what you mean by hitstun multiplier, and hitstun scaled with knockback in the other games as far as I'm aware.

 

[LimitCrown]

Zero Suit Samus's paralyzer has the paralysis effect, which is different from hitstun.

 

[Shaya]

It's a hitbox 'element' though. There may be some various magics to those elements.

 

[LimitCrown]

In regards to stale-move negation, how are attacks that hit multiple times treated? Does each hit count toward the attack being added to the queue, or are those types of attacks added to the queue only once?

 

[Jaxas]

In regards to stale-move negation, how are attacks that hit multiple times treated. Does each hit count toward the attack being added to the queue, or are those types of attacks added to the queue only once?

Hey look, another excuse to link my old research thread!

Note: This was from 3DS, so while it's unlikely that it's changed in the patch it's technically possible.

 

[LimitCrown]

I guess that multi-hit attacks are counted only once. Also, it seems that stale-move negation does not affect any extra shield damage that certain moves may be able to deal. Besides Shulk's Shield Monado Art, I wonder if the rest of the Monado Arts that affect the amount of damage that Shulk takes and the Wii Fit Trainer's Deep Breathing move affect the extra shield damage.

 

[Fox Is Openly Deceptive]

Hey look, another excuse to link my old research thread!

Note: This was from 3DS, so while it's unlikely that it's changed in the patch it's technically possible.

I wasn't aware of this thread, but a little while ago I did some testing of my own and I found a couple of things that aren't in that thread which you may want to look into. http://smashboards.com/threads/mu-link-vs-luigi.394297/#post-18735328
(You may also want to look at the posts further up for context.)

 

[LimitCrown]

If Luigi used only 8 other attacks before trying to fire another fireball, then the move would still be somewhat stale because the move would be at the last part of the queue. Because the queue is the last nine moves that a character has connected with, then 9 other attacks would be needed in order to make the move fresh.

Regarding how Luigi's up-tilt doesn't stale when it is used repeatedly as fast as possible, it may be due to how you can use the same attack again during some of its ending frames. The same thing happens with Little Mac's up tilt and Sheik's side tilt. A similar thing happened in Brawl, it seems.

 

[TheReflexWonder]

I think it's connected to IASA frames earlier than the animation end, yeah. I know that used to happen with pummels in Brawl.

 

[Fox Is Openly Deceptive]

That's interesting. That's why I wanted it looked into further. Either way, it's new information that wasn't in the thread.

 

[thehard]

Weird question- if you hold a direction while being grabbed, do you DI ASAP when thrown?

 

[HeroMystic]

Weird question- if you hold a direction while being grabbed, do you DI ASAP when thrown?

Yes. If you're already holding a direction when you're hit/thrown, you've already successfully DI'd (this is usually why randomly bad DI exists).

 

[LimitCrown]

It appears that if you use any strong up attack or strong side attack in rapid succession, then the attack won't become stale and the series of attacks will be counted only once in the queue. The exception to this is Meta Knight's strong side attack because each part of the attack is counted separately.

Are the damage multipliers for stale-move negation exactly the same as they were in Brawl? It might not be the case because if it was, then Charizard's pummel when it is completely stale would deal 0.92% damage.

 

[Mr. Potatobadger]

Can someone explain Stale moves to me? I'm sort of confused on how the whole thing works. I have some specific questions, though.

During play, it's incredibly noticeable that if you spam an attack, it has less knockback, but unlike what most people say, it doesn't seem to reduce damage (from what I've noticed).

Do stale moves recent when you lose a stock? Can moves become stale if you don't hit anything, or do they only stale when you hit someone/something? (or vice versa)

How exactly do moves regain there power? Can they even regain power or is it permanent? Is it time based? Does using other moves help?

If stale moves do affect how much damage, how is that percent determined? I'm assuming it would be scaled, because a move like fox's side b wouldn't do any damage after a while if it were a set value, whereas a move like Marth's Tippered UpSmash would have barely any deterioration. Unless it were a move specific value? I dunno.

I apologize if my questions are kind of dumb.

 

[Balgorxz]

damage and knockback are reduced, just spam fireballs and you will notice

 

[LimitCrown]

The damage of the attack is reduced, which indirectly reduces knockback as a result. The variable for the damage dealt by an attack in the knockback formula is affected by this, too, which is how it was in Brawl. In Melee, that variable ignored any reduction in damage caused by stale-move negation.

Moves don't stale if they don't hit anything or if they hit an invincible target. For the attacks to regain their power, all that you need to do is to use other attacks. The game records the last nine moves that hit an opponent. For attacks that have multiple parts like most jabs and Marth's Dancing Blade, it appears that each part of the attack is considered to be separate from each other.

 

[Xygonn]

I have a question about attack attributes. I came up on the topic because if you look at the marth thread the rankings for vertical KO are as follows:


Character%
King Dedede204%
Donkey Kong194%
Captain Falcon185%
Ike184%
Ganondorf183%
Bowser183%
Bowser Jr.181%
Mega Man180%
Link176%
Lucario175%
Shulk174%
Wario173%
Samus173%
R.O.B.173%
Duck Hunt Duo171%
Charizard171%
Diddy Kong169%
Pit168%
Marth168%
Lucina168%
Little Mac168%
Dark Pit168%
Robin167%
Mario167%
Dr. Mario167%
Yoshi165%
Sonic164%
Pac-Man163%
Villager161%
Luigi160%
Zero Suit Samus159%
Wii Fit Trainer159%
Ness158%
Greninja158%
Falco158%
Toon Link156%
Zelda155%
Palutena155%
Sheik154%
Peach154%
Fox153%
Meta Knight152%
Olimar149%
Rosalina % Luma148%
Pikachu148%
Kirby145%
Mr. Game & Watch138%
Jigglypuff134%

But if you look at the Shulk Usmash rankings from the weight thread the order is slightly different:

Vertical:
1. Shulk (with Shield Minado Art) - 120%

2. King Dedede - 85%

3. Donkey Kong - 82%

4. Bowser - 80%

5. Bowser Jr., Ganondorf, Ike, Captain Falcon, Megaman - 76%

10. Wario, Link, Charizard, R.O.B. - 74%

14. Lucario, Shulk (no Minado Arts) - 72%

16. Yoshi, Samus, Duck Hunt - 71%

19. Mario, Diddy Kong, Dr. Mario - 70%

22. Pit, Marth, Robin, Dark Pit, Lucina - 69%

27. Villager, Pac-man, Sonic - 68%

30. Luigi, Little Mac - 67%

32. Toon Link, Ness, Wii Fit Trainer - 66%

35. Greninja - 65%

36. Palutena, Falco - 64%

38. Peach, Zelda, Zero Suit Samus - 63%

41. Sheik, Meta Knight - 62%

43. Fox, Pikachu, Olimar, Shulk (Smash Minado Art) - 61%

47. Rosalina, Kirby - 58%

49. Mr. Game & Watch - 56%

50. Jigglypuff - 51%

What sticks out to me is that Yoshi is the same as Samus in the Shulk list (rank 16), but is below Mario in the Marth list (rank 26). There are other minor differences, but they could matter too. Is it possible Yoshi is somehow resistant to "pierce"? Does this affect things that are "slash", "fire", "electric", "ice", "water", "darkness", etc.? I couldn't find anything on the attributes of attacks actually mattering with respect to knockback or damage per character. Has this been looked at before?

 

[LimitCrown]

It seems that when an attack fully stale, the amount of damage dealt seems to be approximately 0.528 times the base value. It means that stale-move negation may function slightly differently. The multiplier for fresh moves seems to be the same as it was in Brawl, though.