Momentum conservation

[Rayzorium]

I was surprised to find that no extensive testing has been done on this yet (and if there has, I'm surprised by my own stupidity). I'm talking about how when you're hit by two moves in succession, the momentum from the first attack sometimes affects the angle at which the second attack launches you. Does anyone know how it works?

For example, in a Sheik-only HRC strat, the momentum of the sandbag bouncing off the back wall is conserved when you hit it with the bat (Sheik normally hits it almost straight up, like Falcon), so it is launched at a nice angle. However, in one variation of Ganon's HRC strat, the bat knocks the bag into Warlock Punch, which sends it the other way and completely ignores the knockback from the bat throw.

Another example most people would be more familiar with: Falco's shine-dair. As you know, if a Falco sends you off the stage with a shine immediately followed by a dair, the knockback from the dair is often heavily stifled. Its downward knockback is neutered by the shine's upward launch. However, sometimes it seems to not be affected at all and knocks them down at full strength.

Edit: Results from my testing are below in post #7.

 

[CloneHat]

I thought this too, but the knockback from Falco's shine-dair is not stifled, the player is either inputting the dair too quickly, which causes a weaker hit to connect, or simply hitting the opponent at low percents.

If you are rested by Jiggs, then quickly jabbed by another player, the rest's KB will be canceled and you will only feel the effects of the jab's KB. In Brawl this in not the case, but in Melee, any hitbox with knockback will cancel a character's momentum.

 

[Rayzorium]

I think I may have explained poorly, as the difference in power should be completely unmistakable. No part of the dair could ever cause the "stifled" look without shine preceding it. I just tested it in training mode on FD, and CPU Ganon could actually get back to the stage after being hit with a shine-dair at 130%, but wouldn't even get a chance to use his second jump before dying if daired normally at 100%. However, sometimes, for whatever reason, the spike in shine-dair would be full strength and kill Ganon easily.

Another thing came to mind, actually, when you mentioned jabbing someone to save them from a rest. If you knee someone with Falcon, and jab immediately (low percents, of course), they will regain control due to jab's pathetic hitstun, but usually continue moving backwards from the knee's knockback. I say "usually" because the jab sometimes stops them completely, as you described. I have an idea as to why these inconsistencies happen but I hesitate to share it until I'm sure the issue is understood.

Also, I mentioned this in my first post, but perhaps didn't stress it enough:
http://www.youtube.com/watch?v=fP7LCP3YXdE

Sheik's bat swing normally knocks the sandbag almost straight up, resulting in very, very little distance. In the video, however, the sandbag's trajectory is clearly affected by its flight path prior to getting hit. Coincidentally, Sheik's first three pairs of hits on the sandbag are evidence of complete momentum cancellation. The strategy wouldn't work without both cancellation and conservation not only coexisting, but behaving consistently.

Blarg. If this were a few years ago, I'd be able to just try to figure it all out myself. Help me out here, at least with ideas. Can you think of any other situations where knockback is not completely cancelled between moves?

 

[mers]

This discussion has actually been going on in a few places recently. It seems that certain moves, most notably Fox and Falco's aerial shines, don't entirely override the previous momentum. What you're saying about Falco's shine-dair indicates that this may also apply to moves immediately following the shine, which is very interesting.

One very tentative guess is that it has to do with how quickly the first move hits after the second move. Obviously this is not a complete explanation, as Falco's lasers will obviously stop any momentum immediately, as anyone who has ever played teams with one knows. However, it definitely happens, and I would love to hear an explanation, or at least know what situations this occurs in.

 

[Kason Birdman]

Falcos lazers (and commonly jabs) stop all momentum from the previous hit.. but I find that some other moves dont.

we all know that teching doesnt cancel momentum.
http://www.youtube.com/watch?v=GOg1MIPtwsQ

 

[Fortress | Sveet]

shine has a property in the air such that it doesn't override the previous momentum. theres never been proof that other moves are like this, but i have a hunch a few other character's b moves work this way too

 

[Rayzorium]

O.o

I just checked it out, and I don't think there's anything at all unique about jabs, shines or Falco's laser in this context. It all seems to boil down to what ReverendGadgetBoy said about the importance of how quickly one move comes after the other.

I did some testing in training mode on FD against a target set to moderate %. Using a bob-omb for the initial hit, I found that if the second hit comes sufficiently early (i.e. well before traversing a quarter of the stage), the momentum from the first hit is completely cancelled. Even Fox and Falco's aerial shines sharply interrupted the bob-omb's knockback and behaved as if they were used on a stationary target.

However, if I perform the second hit at a greater distance (about halfway across the stage or more), the momentum from the first hit always affects the second hit. Falco's laser not only failed to completely halt knockback, it seemed to have no effect at all. It brought them out of stun, but they still flew unimpeded to their deaths.

So yeah, the behavior is actually very consistent and predictable. Shine and Falco's laser are no different from any other move, and and every move seems to be capable of either cancelling or conserving momentum depending on timing. The only factor determining whether or not it happens is how quickly the second hit follows the first.

Edit2: Also, it seems that same-directional vectors can't add very well (although opposite ones subtract just fine). If you get hit by two moves of similar power in the same direction, and momentum is conserved, the second hit won't be any stronger from the combination (ignoring the higher percentage).

 

[CloneHat]

Oh wow that's really interesting.

I'm gonna test this myself.

 

[Rappster]

dear magus420, have you seen this particular thread yet?

 

[mastermoo420]

O.o

I just checked it out, and I don't think there's anything at all unique about jabs, shines or Falco's laser in this context. It all seems to boil down to what ReverendGadgetBoy said about the importance of how quickly one move comes after the other.

However, if I perform the second hit at a greater distance (about halfway across the stage or more), the momentum from the first hit always affects the second hit. Falco's laser not only failed to completely halt knockback, it seemed to have no effect at all. It brought them out of stun, but they still flew unimpeded to their deaths.

Just imagining this situation with SSBM physics in mind, this looks completely right, lol.

 

[ShadyMilkman]

Did anyone else get entirely sidetracked watching Sin's insane collection of HRC videos?

Everytime I stumble upon one on Youtube, I always wind up staying for at least 20-30 minutes.

 

[A Pimp Named Slickback]

you should make videos or have others make a video explaining this.

then edit your OP with results, conclusion, and recorded evidence.

then get everybody to suck your ****.

or something I dunno.

 

[FoxLisk]

so. i was thinking about this the other day, and it reminds me of something i was shown back in the early days of melee. like, '01 ****.

if you go to great bay (the link stage with the turtle), and have a peach and another character, and stand on the part of the platform that goes under the top platform, you can observe something similar. use Peach's downtilt like, ten times. the other char will have a **** ton of momentum. you'll be in that absurd floating animation for a long time, with your movement blocked by the top platform. you can slide along it and everything. if you slide so far that the platform isn't blocking you anymore, you'll fly up the top.

i think this is relevant, although it might just be a different phenomenon.

 

[nube]

so. i was thinking about this the other day, and it reminds me of something i was shown back in the early days of melee. like, '01 ****.

if you go to great bay (the link stage with the turtle), and have a peach and another character, and stand on the part of the platform that goes under the top platform, you can observe something similar. use Peach's downtilt like, ten times. the other char will have a **** ton of momentum. you'll be in that absurd floating animation for a long time, with your movement blocked by the top platform. you can slide along it and everything. if you slide so far that the platform isn't blocking you anymore, you'll fly up the top.

i think this is relevant, although it might just be a different phenomenon.

http://www.youtube.com/watch?v=FUqZ3E6qRsc#t=45s

 

[FoxLisk]

http://www.youtube.com/watch?v=FUqZ3E6qRsc#t=45s

haha. yes! that is beautiful. given the absurd percents, however, it doesn't make it quite as obvious that momentum is stacking.

 

[StoneColdBeast]

*to the last couple of posts*, that's nothing weird, momentum is ALWAYS conserved when you're sent flying in a "ready" (non-tumbling) position, that's not the point of this topic. The point of this topic is the strange case when you're hit by a hitbox and momentum of the character before the hit plays a role in the final trajectory.


That being said, I approve of this thread, it's very interesting. Nobody's gotten ahold of Magus? I think this thread isn't popular enough >_<


also, lol, this thread reminds me a lot of physics, especially when symmetries are spontaneously broken in particle physics..


But the theory that it has to do with how fast the second move is done does sound right, I'm just interested in the details.

 

[exarch]

too a point the second move is true... but it seems when it happens too late it goes back to overriding.
http://www.youtube.com/watch?v=GOg1MIPtwsQ#t=41s
Although it's possible that Ganon's momentum vs the second hit is too little so we don't really notice it.

 

[Rayzorium]

Unfortunately, I don't really know what details there are to get at - possibly narrowing down the timing window, or uncovering the reason for it, but I don't know if that's within my ability. C'mon, Magus! I haven't really been around but I'm guessing he's the man for game mechanics these days?

Also, regarding going back to overriding, exarch, I think it's nearly impossible to tell just from the video - as you said, Ganon had already lost some momentum by then, and I wouldn't really know how to interpret turtle knockback anyway.

I did do a little test on Hyrule, however - I set Falcon to 999%, d-tilted him with Peach to "stick" him to the roof, set him to 0%, and glided him to an upward opening where a proxy mine was waiting. Even with a several second gap, the mine ended up killing him, at 0%, off the top. I even repeated it a few times, increasing the time gap until it no longer killed Falcon off the top (but still sent him much farther than a proxy mine should at 0%). No matter how late the proxy mine hit came, until Falcon completely lost all energy from Peach's d-tilt, the explosion always sent him farther than expected.

 

[A Pimp Named Slickback]

it's possible that it has something to do with the stages of hitstun

i know that in Brawl, when you get there are several stages of hitstun. the initial hitstun, which you can't airdodge and stuff out of (quite a small amount), which is followed by a second stage which you can airdodge and attack out of.

I think it's something like that except without any of the idiotic airdodge garbage.

 

[Pen_island]

So Im not the only one that noticed that shine>dair behaves strangely sometimes

 

[BigD!!!]

slickback, the "second stage" of hitstun youre referring to is more commonly described as "not being in hitstun anymore"

 

[AXE 09]

O.o

I just checked it out, and I don't think there's anything at all unique about jabs, shines or Falco's laser in this context. It all seems to boil down to what ReverendGadgetBoy said about the importance of how quickly one move comes after the other.

I did some testing in training mode on FD against a target set to moderate %. Using a bob-omb for the initial hit, I found that if the second hit comes sufficiently early (i.e. well before traversing a quarter of the stage), the momentum from the first hit is completely cancelled. Even Fox and Falco's aerial shines sharply interrupted the bob-omb's knockback and behaved as if they were used on a stationary target.

However, if I perform the second hit at a greater distance (about halfway across the stage or more), the momentum from the first hit always affects the second hit. Falco's laser not only failed to completely halt knockback, it seemed to have no effect at all. It brought them out of stun, but they still flew unimpeded to their deaths.

So yeah, the behavior is actually very consistent and predictable. Shine and Falco's laser are no different from any other move, and and every move seems to be capable of either cancelling or conserving momentum depending on timing. The only factor determining whether or not it happens is how quickly the second hit follows the first.

I have a theory that there's a "sweet spot" DI on the second hit to negate any influence from the first, but I only have one controller, so it's hard for me to test. Just a theory.

Edit: Also, it seems pretty clear that vectors can't add, only subtract. In other words, energy from the first hit can slow the second hit down, or change its angle, but will never make it stronger or faster.

Thank you so much! Everything makes a lot more sense now. Seriously, thank you for clearing things up. I've been wondering about this for a while, but we still don't know EXACTLY how it works. Like... what determines whether the second hit will just send you flying normally, or only change the trajectory of the 1st hit? I mean... how much time EXACTLY does the second hit have to be delayed in order to get that "stifled" effect?

 

[oakleaf]

how much time EXACTLY does the second hit have to be delayed in order to get that "stifled" effect?

We need some kind of mew2king statistics tables, or some Magus advanced gif analytics

 

[Winston]

I just checked it out, and I don't think there's anything at all unique about jabs, shines or Falco's laser in this context. It all seems to boil down to what ReverendGadgetBoy said about the importance of how quickly one move comes after the other.

My bad about spreading this idea I guess >.>

I was arguing about why doubleshine works in the fox boards and some people didn't think momentum was ever conserved so I made this video:

http://www.youtube.com/watch?v=zsMK-TJuRIg

which led me to believe it was a special property of the shine. I was dumb and didn't test it with other moves like jabs and etc. though. Aerial shine probably behaves like every other move then.

I still think that grounded shine might have the special property of always canceling momentum though, since I don't know how to explain the difference in that video then. (I got the same result consistently, so I don't think it was differences in timing.) Did you do your bo-bomb tests with aerial or grounded shines?

Also, I think your theory about the timing might mean that a perfectly executed doubleshine might be worse than a slower doubleshine for gimping, since a perfectly executed one might cancel all the existing momentum.

 

[mers]

Also, I think your theory about the timing might mean that a perfectly executed doubleshine might be worse than a slower doubleshine for gimping, since a perfectly executed one might cancel all the existing momentum.

Nope. Every shine except the first one you do while multishining technically classifies as an aerial shine, even if you do it perfectly. That's the reason you have to time it frame perfectly, as opposed to things like JC grabbing/usmashing. You shine on the first frame Fox/Falco leaves the ground, so the shine is aerial, you just land instantly (probably the next frame but I'm not entirely sure about that).

So, no worries there.

 

[smakis]

pretty sure above is wrong, landing doesnt cancel the shine, jumping does

you shine, jump then cancel the jump before you leave the ground with another shine, even if you landed on the next frame you would still have landing lag and lag from releasing the shine

 

[MooseEatsBear]

pretty sure above is wrong, landing doesnt cancel the shine, jumping does

you shine, jump then cancel the jump before you leave the ground with another shine, even if you landed on the next frame you would still have landing lag and lag from releasing the shine

That's wrong, the post above yours is right.
http://www.smashboards.com/showthread.php?t=214549
You shine, then on the first airborne frame, you shine, which cancels THE HEIGHT of your jump, so you fall back down the slightest bit, which makes it so you land on the ground again and re-gain your initial jump and jump cancel the shine you did while aerial.

 

[smakis]

thats why i said i was pretty sure and not completly =)

 

[Winston]

Nope. Every shine except the first one you do while multishining technically classifies as an aerial shine, even if you do it perfectly. That's the reason you have to time it frame perfectly, as opposed to things like JC grabbing/usmashing. You shine on the first frame Fox/Falco leaves the ground, so the shine is aerial, you just land instantly (probably the next frame but I'm not entirely sure about that).

So, no worries there.

yeah, I know.

So, under the old theory where aerial shine always preserves momentum, there is no difference.

however Rayzorium has found that aerial shine sometimes cancels momentum; it just depends on the timing.

So my point is that a perfectly executed doubleshine happens faster and therefore might cancel the momentum due to the second shine happening very soon after the first. It's a possibility.

 

[Gea]

Are you sure the momentum isn't just added unless the momentum is below a certain threshhold/moves with "no" knockback are used?

Like what if Falco daired that Falcon as he was pinned to the roof and you put him back at 0%? The momentums are nearly (if not) opposite, but the hit downwards should be much weaker than the current momentum, so it should just slow him down if I am reading this correctly. When I get home I will test this out.

 

[TheCrimsonBlur]

Jesus christ so many theories. This seems so fundamental too.

Someone contact Magus already

 

[Rayzorium]

Whoa, lots of replies. I'll just go in order of the thread.

it's possible that it has something to do with the stages of hitstun...

... how much time EXACTLY does the second hit have to be delayed in order to get that "stifled" effect?

Stages of hitstun... the thought crossed my mind, but it would be pure speculation. Still, I'm not terribly comfortable with calling it a matter of "timing"; I just said that because I wasn't sure how else to put it. I think anything exact would be difficult to find without AR (or M2K, I guess, but I don't think he'd be interested >_>).

Did you do your bo-bomb tests with aerial or grounded shines?

I made sure to use both aerial and ground shines because I had heard about aerial shines being special, and found that there was no difference. I think the reason you got consistent results was that you always jumped into aerial shine and thus would hit Bowser slightly later than you would with standing shine. Just now, I got both outcomes with both ground and aerial shines. Serendipitously, your setup seems to put us right on the cusp of the timing window and is very easy to reproduce!

I don't know what you mean exactly by doubleshine gimp, but I don't think the timing window is so tight that a slow doubleshine would be any different from a perfect one. Your red shell setup goes a long way toward getting a feel for the timing too . However, I don't think the momentum stacks, exactly. Same-directional vectors don't add very well (although opposite ones subtract just fine). A second attack that sends you the same general direction as the first wouldn't see much change. Again, figuring out the exact interaction would almost certainly require AR.

Are you sure the momentum isn't just added unless the momentum is below a certain threshhold/moves with "no" knockback are used?

I tested quite a few moves, and every move was able to interact in some way, be it Falco's laser or Warlock Punch. Even though you were going to test it, I couldn't help myself and checked it out too. I got off 2 dairs before accidentally letting him slide off the ceiling (as dair has a little horizontal knockback), at which point he flew into space, most of his momentum still intact.

I did notice that shining him instantly unstuck him, but I'm sure this is because it caused him to "bounce" off the ceiling. I think the bounce itself counts as a hit, and because it cameinstantly after shine, it was well within the timing window for cancellation.

Jesus christ so many theories. This seems so fundamental too.

I was really shocked at how simple it turned out be - I had been expecting conservation to be a property of a few particular moves, with various conditions and technicalities. Something this systematic and predictable was a pleasant surprise. I actually don't think there are that many theories, and nothing currently contradicts anything else. It's just a matter of figuring out the details. Maaaaaaaguuuuuuus...

 

[Winston]

Alright, cool. I'm just dumb and didn't test it enough.

I feel really silly now because I told everyone on the fox boards that it was unique to aerial shine and they believed me ><

the doubleshine gimp is this:

http://www.youtube.com/watch?v=Z4nWF5r5VI0

based on what you found though, it seems like the doubleshine shouldn't be any different from the single shine. It's two hits one right after the other, and it's two vectors in the same direction.

So I'm pretty confused now. How do you explain doubleshine being better than single shine for gimping? Or do you think there is no difference?

 

[Gea]

It's better in the sense that the second hit registers when he is lower and towards the edge more, instead of above him. He doesn't seem to travel any farther though, otherwise he wouldn't have been able to make it back via up + B (or try to).

 

[Rayzorium]

Ohhh, that. Actually, even with identical vectors, I can't completely rule out the possibility of differences because I simply don't have the tools, but a delayed doubleshine is tight enough to where it shouldn't be an issue anyway. Doubleshine is still superior, IMO, for other reasons, although my Fox/Falco theory is nothing to write home about.

First, as Gea said, position. Also, I think Falco can normally just barely double jump and airdodge back onto the stage from a single shine. This could be enough to remove that as a possibility.

Second, timing. I think most smashers have an extremely good feel for when they get out of hitstun and can double jump (especially in response to being shined off the stage), and I think doubleshine can upset this. Falco double jumped a little late on both gimps, not, I suspect, because of any mechanical factor, but because of thrown timing.

And I guess it's possible that they might stop DIing prematurely and get wrecked extra hard by the second shine, or get "intimidated" by the mix-up, but I think the first two reasons are far more likely.

Honestly though, I think he stood a good chance of dying even if those had been single shines - that's just a really nasty position for Falco to be in.

 

[Winston]

One reason I thought it sends them farther is that in training mode, the falco falls wayyy farther down before starting his up b if it's a doubleshine. It's a CPU so I'm not sure why it'd mess up the CPU's timing.

Doing it from the top platform though, I can't see any difference in trajectory. So I don't know.

 

[Gea]

CPUs are weird. Best way to test is with AR if you have it.

 

[StoneColdBeast]

I just f-smashed a giant wire frame on battlefield with falco. after flying about half way across the stage an articuno (the pokemon) froze him and he died off the side. articuno kills straight up and there's no way the wire frame could have DIed far enough to die of the side of the stage. Clearly the fsmash momentum combined with articuno ice thing..

 

[X1-12]

thats not the best example, especially considering the physics of the game change when you are fighting wireframes..

 

[Metà]

related video

http://www.youtube.com/watch?v=315AWjxI7ZE

???????????????