theEffinBear
Smash Apprentice
[size=+2]
[/size]
Snake, Kirby, Yoshi,
all have varying degrees of knockback-ignoring moves (Cypher, Stone, double jump,
), but many explanations for how/why/when this happens are contradictory or aren't correct in EVERY situation. So I decided to find out for myself: this topic will be about every knockback-ignoring move that is not Super Armor, starting with Snake's Cypher Armor, but later expanding to include Kirby and hopefully Yoshi too, which is why I have placed it in the Brawl Tactical Discussion forum and not the Snake forum (though I'm linking from there to here); if I have forgetten other non-Super Armor moves, or if more are discovered, then the topic will include them as well.
I think this is information is new, as I have looked through all the threads that seemed relevant on the Snake forums and on the Tactical forums within the first few pages, I have performed some searches, and come up blank for "when EXACTLY will a move interrupt the Cypher?", though the closest thread I found was Snake's up-B - Acts like super armor? back in April. If someone else has completely figured this out before, then sorry: the information was buried deeper than my reasonable effort was able to uncover. Anyway!
Onwards to SCIENCE.
Everyone knows that Snake can be hard to hit off the Cypher: sometimes you hit him and he just powers through it, like Super Armor. Most single-hit moves do break through this Cypher Armor: Kirby's f-tilt always launches him, 7 or 8% the worse for wear, for example, but most multi-hit moves fail to affect his recovery. But then I noticed that Charizard's b-air, a 2-hit move, breaks him out on the second hit. And then I tried Snake's own d-air and n-air on him, and the last hit of both 4-hit moves broke through the Cypher Armor. So I've heard some people say that each move in the game is has a TRUE or FALSE flag marking "INTERRUPTS_CYPHER". But Lucario at 200% will break through the Cypher Armor with either of the hits on his 2-hit d-air, while Lucario's d-air at 0% can only deal damage to a Cypher'ing Snake, no knockback. And the initial hit of Squirtle's n-air deals 12% and interrupts the Cypher, but hitting slightly later deals 6% and doesn't interrupt. Why these inconsistencies and differences?
Turns out, there's one single condition for whether an attack will break through the Cypher Armor: if Snake takes
damage from a single hitbox, he actually takes the knockback instead of ignoring it. The individual hitboxes on multi-hit moves usually don't
that 7% threshold, so their knockback is ignored. On the flipside, most single hit moves DO
the 7% threshold, and so they will knock him off the Cypher. I am pretty confident about most of this, having just tested it fairly extensively; my caveats are 1) I don't know how/if move decay counts in this, as 2) I did this all in Training Mode (it would be a major downer to try setting all the percentages so exactly during a Brawl instead of Training).
Method:
0) choose Snake in training mode with Lucario as your CPU opponent.
1) Use up-B and have Lucario hit you with his d-air, noting the damage taken from each hit and whether either knocked you off the Cypher.
3) Spawn a Maxim Tomato and eat it. Have Lucario hit you with his u-tilt, noting the same things.
4) Spawn another Maxim Tomato and eat it ("Tasty!") and increase Lucario's damage by 10%.
5) Repeat from step 1.
6) When you find that one of these moves knocks you off the Cypher, start narrowing down to the exact percentage that Lucario is powerful enough to knock you off (I didn't test the u-tilt while doing this for the d-air, hence the gap in that column).
7) Test to make sure that a 6% attack doesn't break the Cypher Armor when Snake is very damaged.
8) After all this, switch to Lucario with Snake as the CPU, and make sure the results hold in a second trial with Snake at 999%.
I chose Lucario because his Aura ability provides a very handy way to slowly pump up a single move, testing it with different damage. I chose the d-air and the u-tilt because they start out too weak to interrupt the Cypher, because they're both fast enough that it's relatively easy to test this at 1/2 speed in Training mode while using two controllers, and because the d-air is a multi-hit move.
The table below is the summary of my findings, where "doesn't interrupt" is in blue and "does interrupt" is in red; Lucario's percentages in green are when I was narrowing down in step 6 from above (so they're not in nice even 10% increments).
So Lucario's d-air at 50% does 13% in two hits, neither of which knock you off the Cypher (the first hit does 6% by itself, and if you intentionally miss with the first hit, the second hit does 6% by itself as well); Lucario's d-air at 60% does 14% in two hits and does knock off (both dealing 7% individually), and both of the hits knock you off the Cypher. So test the halfway point, 55%, which deals 7+7=14%, still knocking you off. Halving and truncating the difference yields 52%, which deals 6+6=13%, and doesn't knock you off. Test 53%, which deals 7+7=14%, and does knock you off. So, 53% is the critical point, when both hits of his d-air do 7%
, and both can knock you off of the Cypher.
The same thing happens with the u-tilt, except that 114% deals 6% and doesn't interrupt the Cypher, but at 115% the u-tilt deals 7%
and does interrupt the Cypher.
RESULTS:
• As stated above, if an individual hitbox deals
damage, the Cypher ignores the knockback. If an individual hitbox deals
damage, the Cypher will be interrupted.
• Knockback has no direct effect on whether a hit will break the Cypher Armor, only the indirect correlation effect that more knockback usually means more damage on an individual hitbox.
• Training mode doesn't just ignore fractions of a percent, or else Lucario's d-air at 0% couldn't deal 11% total, as each hit only registers as 5%.
• There's no "INTERRUPTS_CYPHER" flag for each and every single attack in the game.
• Both hits of Lucario's d-air deal exact same amount of damage.
• The knockback immunities of Snake's Cypher, Yoshi's double jump, and Kirby's Stone are all different (unique?) mechanics.
FURTHER STUDY REQUIRED:
• This was done in Training Mode, and it's possible that there are some differences between Training Mode and actual Brawls. Hopefully, the only difference between Training and Brawls are 1) the known lack of Move Decay in Training, and 2) that the display (ONLY the display) truncates your % in Training and rounds it in multiplayer, or something along these lines.
• I don't know whether or how Move Decay interacts with Cypher Armor (I didn't think to test it out before I turned everything off). My GUESS is that if you're Lucario at 115% and you stale your u-tilt at all, it won't break the Cypher Armor. The more-complex alternative is that Cypher Armor doesn't take stale moves into account (so a semi-stale u-tilt at 115% would deal 6%, say, but would still break the Cypher Armor because the FRESH u-tilt WOULD HAVE DEALT
), which is awkward. For Occam's sake, I hope my guess is correct: Lucario's damage-powered Aura seems to be almost the same mechanic as Stale Moves but in reverse, and it would be very strange for the Cypher Armor to take Aura into account but ignore Stale Moves.
• Does Snake have this Cypher Armor for every frame that he's holding onto the Cypher? Does the Cypher Armor end as soon as you regain your ability to airdodge, attack, etc?
So, Dear Readers, this is where I turn to you: does anyone have the patience to check out my three points of Further Study?
• Use the Handicap function to set Lucario at various percentages and test some moves on the Cypher when out of Training Mode: Lucario's 52%-d-air and 114%-u-tilt shouldn't break the Cypher Armor, but both should at 53% and 115%, respectively, if my theory is correct.
• Use the Handicap function to set Lucario to 115% and fight a non-moving Snake on a stage with destructible walls, etc. Stale your u-tilt once by hitting the one of the destructible doodads, and then see if it breaks the Cypher Armor, and how much damage it did (ideally it should do 6% and not break the armor). Repeat the whole thing, increasing Lucario's handicap percent by 1 each time, until the once-staled u-tilt again breaks the Cypher Armor (it should do 7% again as soon as it succeeds in interrupting the Cypher).
• Just test to see if the Cypher Armor applies during the whole Cypher, or if the first few frames are actually Super Armor, or if it the Cypher Armor ends before Snake lets go, etc. My impression is that it's Cypher Armor the whole time, with no actual Super Armor, but I'd like some actual data here.
When you're recording the amount of damage dealt, remember that Snake should have started from 0%, or else you're adding hidden fractions which could mess up the percents you write down (so set items on High, Maxim Tomatoes only, or have Snake commit suicide each time).
Is anyone willing to help me out?
Next up, I think I've figured out how Kirby's Stone Armor works, too, but I'll leave that for a subsequent post (it's not quite as pressing an issue as Snake's Cypher Armor, because the thought "I need to break Kirby's Stone!" doesn't happen quite as often as "I need to interrupt/gimp Snake's recovery!"). I don't know how Yoshi's Jump Armor works, and since the best/most efficient method I can think of to try testing it (using the "fastest launch speed" from the post-Brawl stats, as I don't THINK that the Jump Armor takes damage into account directly) would require all sorts of switches between modes. If anyone has any ideas, or wants to spearhead the Yoshi stuff, please feel more than free. Share and Enjoy.
Note: SmashWiki's explanation for Yoshi's Jump Armor is that he reacts to knockback while double jumping as if his percentage is 120 lower than it displays, and if the resulting calculation is negative, he takes no knockback from that hit. As the same page lists the Cypher as gaining this same exact type of "hit resistance", I would prefer to do reproducible tests with hard data.
Hopefully this is useful to all of you.
/RtEB
.The Armor Project.
.EDIT: I have been updating this post with more details as I find them. Annotated, corrected, and additional info that I have edited in is in this this red-on-black text style, to make clear what is new. I have also changed the name of the topic itself to "The Armor Project"..
.July 23rd: I have completed testing Kirby's Stone Armor. It breaks when it is dealt at least 30% damage (taking Stale Moves into account), all the excess damage from the hitbox that actually broke it disappears, and Kirby takes the knockback (but not the damage) from that hit. Attacking Kirby while he's in his Stone form will Stale the attack, regardless of whether you break the Armor or not. See my [post=4976920]Stone Armor post[/post] below for details on how I tested this..
.August 30th: After a month-long pause, I have tested Yoshi's Jump Armor! It functions against knockback, not damage, which is different from Cypher Armor and Stone Armor. If the attack deals more than 3854±3 units of knockback (based on the move used, how Stale it is, Yoshi's current percentage, and Yoshi's weight), Yoshi's double-jump will be interrupted and he will be sent flying. If the attack deals less than that amount, Yoshi's Jump Armor lets him power straight through the attack, uninterrupted. These knockback values are accessible through the "max launch speed" in the post-Brawl stats, where they are listed in "mph" (unfortunately much less convenient than the damage-based Armors). For testing details and results, see my [post=5253672]Jump Armor post[/post] below..
Snake, Kirby, Yoshi,
.Squirtle (sort of), and Zero-Suit Samus (possibly).
.Withdraw, and down-smash.
I think this is information is new, as I have looked through all the threads that seemed relevant on the Snake forums and on the Tactical forums within the first few pages, I have performed some searches, and come up blank for "when EXACTLY will a move interrupt the Cypher?", though the closest thread I found was Snake's up-B - Acts like super armor? back in April. If someone else has completely figured this out before, then sorry: the information was buried deeper than my reasonable effort was able to uncover. Anyway!
Onwards to SCIENCE.
Everyone knows that Snake can be hard to hit off the Cypher: sometimes you hit him and he just powers through it, like Super Armor. Most single-hit moves do break through this Cypher Armor: Kirby's f-tilt always launches him, 7 or 8% the worse for wear, for example, but most multi-hit moves fail to affect his recovery. But then I noticed that Charizard's b-air, a 2-hit move, breaks him out on the second hit. And then I tried Snake's own d-air and n-air on him, and the last hit of both 4-hit moves broke through the Cypher Armor. So I've heard some people say that each move in the game is has a TRUE or FALSE flag marking "INTERRUPTS_CYPHER". But Lucario at 200% will break through the Cypher Armor with either of the hits on his 2-hit d-air, while Lucario's d-air at 0% can only deal damage to a Cypher'ing Snake, no knockback. And the initial hit of Squirtle's n-air deals 12% and interrupts the Cypher, but hitting slightly later deals 6% and doesn't interrupt. Why these inconsistencies and differences?
Turns out, there's one single condition for whether an attack will break through the Cypher Armor: if Snake takes
.MORE THAN 7%.
.exceed.
.exceed.
Method:
0) choose Snake in training mode with Lucario as your CPU opponent.
1) Use up-B and have Lucario hit you with his d-air, noting the damage taken from each hit and whether either knocked you off the Cypher.
3) Spawn a Maxim Tomato and eat it. Have Lucario hit you with his u-tilt, noting the same things.
4) Spawn another Maxim Tomato and eat it ("Tasty!") and increase Lucario's damage by 10%.
5) Repeat from step 1.
6) When you find that one of these moves knocks you off the Cypher, start narrowing down to the exact percentage that Lucario is powerful enough to knock you off (I didn't test the u-tilt while doing this for the d-air, hence the gap in that column).
7) Test to make sure that a 6% attack doesn't break the Cypher Armor when Snake is very damaged.
8) After all this, switch to Lucario with Snake as the CPU, and make sure the results hold in a second trial with Snake at 999%.
I chose Lucario because his Aura ability provides a very handy way to slowly pump up a single move, testing it with different damage. I chose the d-air and the u-tilt because they start out too weak to interrupt the Cypher, because they're both fast enough that it's relatively easy to test this at 1/2 speed in Training mode while using two controllers, and because the d-air is a multi-hit move.
The table below is the summary of my findings, where "doesn't interrupt" is in blue and "does interrupt" is in red; Lucario's percentages in green are when I was narrowing down in step 6 from above (so they're not in nice even 10% increments).
Code:
Lucario d-air u-tilt
0% [color=#8888ff]5*2=11 4[/color]
10% [color=#8888ff]5*2=11 4[/color]
20% [color=#8888ff]5*2=11 4[/color]
30% [color=#8888ff]6*2=12 4[/color]
40% [color=#8888ff]6*2=12 4[/color]
50% [color=#8888ff]6*2=13 5[/color]
[color=#00ff00]51%[/color] [color=#8888ff]6*2=13[/color]
[color=#00ff00]52%[/color] [color=#8888ff]6*2=13[/color]
[color=#00ff00]53%[/color] [color=#cc4444]7*2=14[/color]
[color=#00ff00]55%[/color] [color=#cc4444]7*2=14[/color]
60% [color=#cc4444]7*2=14[/color] [color=#8888ff]5[/color]
70% [color=#cc4444]7*2=15[/color] [color=#8888ff]5[/color]
80% [color=#cc4444]8*2=16[/color] [color=#8888ff]6[/color]
90% [color=#8888ff]6[/color]
100% [color=#8888ff]6[/color]
110% [color=#8888ff]6[/color]
[color=#00ff00]113%[/color] [color=#8888ff]6[/color]
[color=#00ff00]114%[/color] [color=#8888ff]6[/color]
[color=#00ff00]115%[/color] [color=#cc4444]7[/color]
120% [color=#cc4444]7[/color]
.actually slightly more than 7+fraction%, but I haven't tested enough to know what the fraction is.
The same thing happens with the u-tilt, except that 114% deals 6% and doesn't interrupt the Cypher, but at 115% the u-tilt deals 7%
.actually 7.01%.
RESULTS:
• As stated above, if an individual hitbox deals
.exactly 7% or less.
.more than 7%.
• Knockback has no direct effect on whether a hit will break the Cypher Armor, only the indirect correlation effect that more knockback usually means more damage on an individual hitbox.
• Training mode doesn't just ignore fractions of a percent, or else Lucario's d-air at 0% couldn't deal 11% total, as each hit only registers as 5%.
• There's no "INTERRUPTS_CYPHER" flag for each and every single attack in the game.
• Both hits of Lucario's d-air deal exact same amount of damage.
• The knockback immunities of Snake's Cypher, Yoshi's double jump, and Kirby's Stone are all different (unique?) mechanics.
FURTHER STUDY REQUIRED:
• This was done in Training Mode, and it's possible that there are some differences between Training Mode and actual Brawls. Hopefully, the only difference between Training and Brawls are 1) the known lack of Move Decay in Training, and 2) that the display (ONLY the display) truncates your % in Training and rounds it in multiplayer, or something along these lines.
• I don't know whether or how Move Decay interacts with Cypher Armor (I didn't think to test it out before I turned everything off). My GUESS is that if you're Lucario at 115% and you stale your u-tilt at all, it won't break the Cypher Armor. The more-complex alternative is that Cypher Armor doesn't take stale moves into account (so a semi-stale u-tilt at 115% would deal 6%, say, but would still break the Cypher Armor because the FRESH u-tilt WOULD HAVE DEALT
.7.01%.
.The Cypher Armor reacts to damage after taking Stale Moves into account; see [post=4933987]this post[/post] for details..
So, Dear Readers, this is where I turn to you: does anyone have the patience to check out my three points of Further Study?
• Use the Handicap function to set Lucario at various percentages and test some moves on the Cypher when out of Training Mode: Lucario's 52%-d-air and 114%-u-tilt shouldn't break the Cypher Armor, but both should at 53% and 115%, respectively, if my theory is correct.
• Use the Handicap function to set Lucario to 115% and fight a non-moving Snake on a stage with destructible walls, etc. Stale your u-tilt once by hitting the one of the destructible doodads, and then see if it breaks the Cypher Armor, and how much damage it did (ideally it should do 6% and not break the armor). Repeat the whole thing, increasing Lucario's handicap percent by 1 each time, until the once-staled u-tilt again breaks the Cypher Armor (it should do 7% again as soon as it succeeds in interrupting the Cypher).
.Since the handicap function has a resolution of 10%, I was slightly more creative in testing this. The Cypher Armor does take Stale Moves into account before reacting to damage (breaking for more than 7% damage like usual), as stated in the above edit; continue to see [post=4933987]this post[/post] for details..
When you're recording the amount of damage dealt, remember that Snake should have started from 0%, or else you're adding hidden fractions which could mess up the percents you write down (so set items on High, Maxim Tomatoes only, or have Snake commit suicide each time).
Is anyone willing to help me out?
Next up, I think I've figured out how Kirby's Stone Armor works, too, but I'll leave that for a subsequent post (it's not quite as pressing an issue as Snake's Cypher Armor, because the thought "I need to break Kirby's Stone!" doesn't happen quite as often as "I need to interrupt/gimp Snake's recovery!"). I don't know how Yoshi's Jump Armor works, and since the best/most efficient method I can think of to try testing it (using the "fastest launch speed" from the post-Brawl stats, as I don't THINK that the Jump Armor takes damage into account directly) would require all sorts of switches between modes. If anyone has any ideas, or wants to spearhead the Yoshi stuff, please feel more than free. Share and Enjoy.
Note: SmashWiki's explanation for Yoshi's Jump Armor is that he reacts to knockback while double jumping as if his percentage is 120 lower than it displays, and if the resulting calculation is negative, he takes no knockback from that hit. As the same page lists the Cypher as gaining this same exact type of "hit resistance", I would prefer to do reproducible tests with hard data.
Hopefully this is useful to all of you.
/RtEB