Simple facts that you may not know...

[SSBPete]

ok im really confused

maybe the slope isn't linear?

 

[The Star King]

I agree with your theory Sangoku, I think it's as slightly sloped as can be. I think the Kirby thing might be unrelated.

 

[RickySSB]

The CP doesn't try to recover again when he get hit while recovering the first time.

 

[prisonchild]

The CP doesn't try to recover again when he get hit while recovering the first time.

false to a certain extent...

 

[SSBPete]

its something to do with the position of the cp. i think if they're above the stage then they try to recover a second time.

i think

 

[clubbadubba]

its something to do with the position of the cp. i think if they're above the stage then they try to recover a second time.

i think

This is exactly it. I think

 

[Sangoku]

I wouldn't say exactly, because there are some weird occurence where the CP is a bit above the stage and still doesn't recover.

 

[The Star King]

its something to do with the position of the cp. i think if they're above the stage then they try to recover a second time.

i think

Pretty sure it's just random, like whether the CPUs tech or not. You can hit the CPU the exact same way in Training Mode, and sometimes it'll recover, sometimes it won't. On DK at Hyrule, I tried drilling him for 30% for Kirby, then FSmashed him to the left 3 times in the same way, and one time he recovered and one time he didn't.

 

[clubbadubba]

I wouldn't say exactly, because there are some weird occurence where the CP is a bit above the stage and still doesn't recover.

Vids or it didn't happen. I would believe that hitting a wall might make them recover again, but I don't think so.

@sk, its def not random, at least nnot for lvl 9s. I'm able to predict w 100% accuracy whether cpu's recover, so can't be random (or I'm God)

 

[prisonchild]

it depends on the kind of hit too i think, a recovering DK when hit by fox's laser will stop dead and fall, even if above the stage. but with a luigi (either nair or upb) he will try to recover again, even if he's at stage level (or slightly below? i wasn't really paying attention)

 

[ballin4life]

Pretty sure it's just random, like whether the CPUs tech or not. You can hit the CPU the exact same way in Training Mode, and sometimes it'll recover, sometimes it won't. On DK at Hyrule, I tried drilling him for 30% for Kirby, then FSmashed him to the left 3 times in the same way, and one time he recovered and one time he didn't.

are you TASing this? otherwise you can't say that you're really doing it the EXACT same way.

also the discussion was about level 9s, not training mode CPUs (which have different behavior)

 

[The Star King]

are you TASing this? otherwise you can't say that you're really doing it the EXACT same way.

Yes I can noob, I avoid pushing him and make sure the hitbox reaches first, and that guarantees identical positioning.

also the discussion was about level 9s, not training mode CPUs (which have different behavior)

Fair enough, but are you sure their recovery behavior is different?

 

[SSBPete]

sk is talking about training mode dummies and clubba and i are talking about lvl 9 cps

i think we're all right

 

[clubbadubba]

sk is talking about training mode dummies and clubba and i are talking about lvl 9 cps

i think we're all right

Hooray for correctness! Yea totally missed the traininng mode talk, blame on the phone

 

[ballin4life]

Yes I can noob, I avoid pushing him and make sure the hitbox reaches first, and that guarantees identical positioning.

small differences in the positioning of your character may matter in their algorithms. i doubt you can say you're positioning your own character in exactly the same fashion (down to the pixel).

Fair enough, but are you sure their recovery behavior is different?

they are totally different AIs; why would their recovery be the same? also my anecdotal experience is that their recovery behavior is different (in training mode sometimes they don't upB at all. haven't seen level 9s do that)

i'm pretty sure that level 9s will only attempt to upB again if their horizontal positioning places them above the stage.

 

[The Star King]

small differences in the positioning of your character may matter in their algorithms. i doubt you can say you're positioning your own character in exactly the same fashion (down to the pixel).

OK fine. I didn't move at all with the control stick, and moved myself only with Falcon Punch and Falcon Kick. Method: Falcon Punch once on the ground to go a bit forward, Falcon Punch again, Falcon Kick to get to Mario, Falcon Punch, Falcon Kick to get down in the tent, Falcon Punch, Falcon Punch. Also I made sure Mario never teched, if he teched I reset. There, same positioning.

Sometimes Mario up-b'ed, and sometimes he didn't. ggs scrub

they are totally different AIs; why would their recovery be the same? also my anecdotal experience is that their recovery behavior is different (in training mode sometimes they don't upB at all. haven't seen level 9s do that)

I tried BThrowing lvl 9 Mario like 30 times and he tried to up-b every time. So I guess you're right. I'm still right about the training mode CPUs though. And Pete is right about everyone being right, which means he's included in the rightness. wut.

Actually I could not get Mario to not up-b at all, even when I baired him repeatedly out of it. So I think it's character dependent. Mario seems to always try.

i'm pretty sure that level 9s will only attempt to upB again if their horizontal positioning places them above the stage.

Wrong, for Mario anyways, I just saw Mario up-b well below the stage (on Dreamland). But maybe that's just Mario.

 

[ballin4life]

OK fine. I didn't move at all with the control stick, and moved myself only with Falcon Punch and Falcon Kick. Method: Falcon Punch once on the ground to go a bit forward, Falcon Punch again, Falcon Kick to get to Mario, Falcon Punch, Falcon Kick to get down in the tent, Falcon Punch, Falcon Punch. Also I made sure Mario never teched, if he teched I reset. There, same positioning.

Sometimes Mario up-b'ed, and sometimes he didn't. ggs scrub

Are you sure you did each of those inputs on the exact same frames? Characters do sway slightly back and forth when they stand still, right?

I tried BThrowing lvl 9 Mario like 30 times and he tried to up-b every time. So I guess you're right. I'm still right about the training mode CPUs though. And Pete is right about everyone being right, which means he's included in the rightness. wut.

Actually I could not get Mario to not up-b at all, even when I baired him repeatedly out of it. So I think it's character dependent. Mario seems to always try.



Wrong, for Mario anyways, I just saw Mario up-b well below the stage (on Dreamland). But maybe that's just Mario.

I just tested on level 9 Mario and he acted exactly as I said.

I think you misunderstood what I was saying. Level 9s will attempt to use at least one upB. Then, if you hit them while they are still in the air from that upB, they will attempt another upB only when their x position (horizontal position) is in line with the stage (i.e. if they were to fall straight down they would hit the stage).

|aaaaaaaaa|
--------------
pretend the ------- is the stage. they will upB again if they are within the vertical lines "|". That's what I mean.

 

[prisonchild]

also, if you get doug hugged off the stage they just fall after they release you. weird stuff.

 

[chuckj]

i think its just like if you hit him during the upB or after he finished the upB, in the second case, he tries to upB again. i dont think his position really matters :s lvl 9 enemies

 

[ballin4life]

i think its just like if you hit him during the upB or after he finished the upB, in the second case, he tries to upB again. i dont think his position really matters :s lvl 9 enemies

no ... when i tested i always waited until his upB was 100% complete (I can't hit mario out of his upB for ****) and he never tried to get back when he was over the abyss.

 

[chuckj]

hum i dont know, i never have problems to know when the cpu will try to recover again

 

[SuPeRbOoM]

is this really being discussed? >_>

 

[Sangoku]

^Exactly what I thought.

 

[The Star King]

Hey man discussing pointless things is the purpose of this thread, that's why I love it. Sangoku I did not expect that from you

Are you sure you did each of those inputs on the exact same frames? Characters do sway slightly back and forth when they stand still, right?

.

I just tested on level 9 Mario and he acted exactly as I said.

I think you misunderstood what I was saying. Level 9s will attempt to use at least one upB. Then, if you hit them while they are still in the air from that upB, they will attempt another upB only when their x position (horizontal position) is in line with the stage (i.e. if they were to fall straight down they would hit the stage).

|aaaaaaaaa|
--------------
pretend the ------- is the stage. they will upB again if they are within the vertical lines "|". That's what I mean.

OK, I misunderstood. I'll try testing it later.

 

[Sangoku]

I like useless stuff, but for some reason I hate talking about CPU's behaviour.

 

[clubbadubba]

I like useless stuff, but for some reason I hate talking about CPU's behaviour.

Cpu behavior is best! Study it, then maybe you can design a better cpu player. Any chance of that ever happeninng btw?

 

[ballin4life]

btw SK even though I'm somewhat joking about the training mode thing, I still doubt that it's truly random whether they upB or not. AFAIK typically in games the only source of supposed randomness is the player, hence in TASes they can manipulate supposedly random things so that they always get really lucky (e.g. in an Ocarina of Time TAS the enemies/grass will just always happen to drop bombs, even though it's supposed to be random). If games contained truly random events then TASes wouldn't be able to do that.

That's based on my (limited) knowledge of TASes though so I may be wrong.

 

[Sangoku]

I think it depends. If an event is truly pseudo-random, you'll still have to be careful when TASing I guess. If you're loading at a point where the RNG pattern is already determined, then TASing is pointless. We all know (we suppose so at least) that items appear randomly right? When TASing, it has occurred that the same item would always appear at the same time and place, even though I kept loading a previous savestate. As I said, to change the random outcome, maybe I'd have to change something before.

Also low level CPUs act like training mode: they sometimes don't upb. I think that is random (or maybe depending on stuff such as percentage, but always with a random factor) as a way to nerf them (which is stupid). I think the way they are designed is act like a lvl 9 for a small amount of time, then enter in passiveness (walking). Or something a bit like that. So if they happen to be in their "strong" part during their recoveries, they'll upb. This is a pure assumption, nothing verified at all.

Clubba: there has been several posts mentioning better CPU. The last one was a full thread by cmu6eh. Unfortunately we haven't heard of it and I doubt anything will ever happen. Someone (okaygo) also said he and antd would probably be able to reprogram the CPUs with their memory knowledge.

Obviously it takes more than just knowing perfectly how they act to improve them. I for one don't have the knowledge to change anything about the ROM.

 

[SSBPete]

computers can never produce truly random events

Spoiler

but its actually true

 

[The Star King]

btw SK even though I'm somewhat joking about the training mode thing, I still doubt that it's truly random whether they upB or not. AFAIK typically in games the only source of supposed randomness is the player, hence in TASes they can manipulate supposedly random things so that they always get really lucky (e.g. in an Ocarina of Time TAS the enemies/grass will just always happen to drop bombs, even though it's supposed to be random). If games contained truly random events then TASes wouldn't be able to do that.

That's based on my (limited) knowledge of TASes though so I may be wrong.

Obviously nothing is truly random, RNGs are used in video games. It should go without saying that when I say something's "random", I'm saying that it's dependent on the RNG. Not sure why you're bringing it up. I'd say you're playing semantics, but I guess judging by the tone of your post you just don't know much about how computers deal with random events.

EDIT: Your post, and then Pete's post saying what it said as if it's something new... I guess it isn't as common knowledge as I thought. Bleh.

 

[Sangoku]

Lol I thought it was common knowledge too, don't know why Pete said it like that. Also I first said pseudo-random, then I assumed everybody knew when we say "random" we mean "pseudo-random".

 

[RickySSB]

No.........

 

[chuckj]

it was not me, sangoku

 

[Sangoku]

Yeah it was okaygo (just checked). I have no idea why I got it all wrong (maybe I said you because you had a k in your nick and all non-caps letters and KESHIKI because he used to post often in ant-d's threads). Still dumb mistake from me, I'm sorry. Rofl never again am I going to post without checking first, especially since it was easy to do (I knew exactly what thread to look into).

 

[asianaussie]

can you reliably predict the results of a random system in the short run (say, the first 10 digits it throws up)?

if not, then it's random as it needs to be

 

[SSBPete]

DID NO ONE SEE MY TROLL FACE?!!

I WAS CLEARLY JOKING

 

[The Star King]

I don't understand how stating a fact is a joke.

 

[SSBPete]

i was playfully pointing out the obvious

you humorless fiend

 

[The Star King]

You put "but it's actually true", you clearly thought it wasn't obvious. I don't know why you're trying to save face. It's not a big deal. Let's move on.

 

[asianaussie]

hey sk

you're a dirk

Spoiler

but it's actually true