The Judgement Hammer's numbers from most likely to appear to least likely

[Mr.Lemon]

The following list is all of the numbers for G&W's side b. They are in order from least likely to most likely to appear during a match. To figure this out I used G&W side b 384 times and found the average for each of his numbers, I could've gone longer to get more accurate percentages, but 384 times will give us a good reference point to start at.

So starting from the
LEAST LIKELY to APPEAR

Number 3: 9.11%
Number 2: 9.63%
Number 1: 9.89%
Number 7: 10.15%
Number 4: 10.93%
Number 5: 11.19%
Number 9: 11.45%
Number 6 and 8: 13.8%

MOST LIKELY to APPEAR

NOTE* Since the percentages for some of these number are really close they could easily move up or down a place or two if someone were to test this by doing more side b's. But again this is somewhat of an estimate and is NOT definite.

The Range (subtract the largest percent from the smallest)
Range = 4.69%
Numbers 6 and 8 have almost a 5% better chance of appearing then number 3

Conclusion:
So this all shows that even though these percentages seem to be close together and probably will not make a difference, in reality they could change the outcome of a game. Say a 6 came up while G&W side b someone that was off of the stage rather than a 3 all because the 6 has a better chance of appearing. Also, these percentages are not all the same...meaning....they can't be random

 

[ryuu seika]

It has already been proven that nothing within Melee is truly random via use of emulators.
If you save state, load, start a battle, wait an exact number of frames, then pull a turnip (for example), you will get the same result every time.
If, however, any player does something different in the match or chooses a different character, a different time is waited for after the match starts, a different amount of time is spent on the select screens or a different (even seemingly identical) save state is loaded, the outcome will be different. The one thing that does not affect the RNG is your costume choice.

This knowledge cannot be put into play in the real game due to the facts that it requires very precise timing and understanding, it can be messed up by your opponents actions and the fact that some outside influence (console time?) changes it.

 

[Mr.Lemon]

Really? I actually thought of that but wasn't sure if stuff like it was ever proven. I did however do this on all of the neutral stages to see if that would change anything and all of the statistics were either similar or close to it.

 

[ryuu seika]

I don't think there was ever any real proof that the testing was real but there was testing that "proved" it here.

This would suggest that, were you well versed in the workings of the RNG, it would indeed be possible to start each battle with a Judgement 9 by manipulating the starting state to your advantage. After that, your opponent's actions would make it impossible to abuse further but a free KO at the start like that would surely be enough of an advantage.

I find it interesting that 1, the only number you can't get on your first hit, isn't the least likely overall.
Also, from what I understand, you can never hit the same number twice before you cycle through them all, which may skew your results towards the numbers being equal when they are actually even further apart.

 

[Mr.Lemon]

Yea I read up on how you can't get a 1 or a 2 your first time and you can't get a 2 on your second side b either. I kept those in mind but somehow 3 managed to come up the least.

As for Armada knowing how to pull a bomb or stitch face on the correct frame thats pretty cool. But do you think the same mechanics go for G&W's side b? Because even in battle I feel like 6's and 8's do appear more then 1's,2's,and 3's.

 

[ryuu seika]

Seikend's posts on page 2 of that thread onwards show just how it works with turnips and, from what I understand, it's more or less the same for all RNG based moves.

 

[Mr.Lemon]

......Yea I guess...idk I'm still hesitant about it all. I mean it seems that it can only be predicted for the first few milliseconds/frames so what about the rest of the match. I know you said Armada knows how to pull a stitch face whenever he wants but it's sooooooo difficult to time wouldn't it be easier to just figure out the probability of the numbers of G&W's hammer or Peaches turnips?

 

[DerfMidWest]

Armada is a robot. if there were a G&W robot like armada then there would just be 9 hammers all day.

 

[ryuu seika]

I still have the suspicion that you can subconsciously learn what makes a good time to use hammers and the like.

Do people who have played G&W longer tend to get better hammers on average than relative newcomers to the character?

I doubt it would be possible to "learn" it well enough to get nines every time but if Armada really does all that, who knows.

 

[Mr.Lemon]

Yea I mean who knows. I haven't seen a ton on G&W's videos but there are a few nines here and there and of course they are all over the combo videos. I'll keep my eyes open for when they come up and try to figure out the pattern.

 

[DerfMidWest]

I've noticed that they do 9 allot. its this game getting into your head. Smash will be the end of the world.

 

[ryuu seika]

Combo vids don't count. They only show the good bits so the proportion of nines goes way up.

@Aesir: The phrase is apparently "emanu elish". The death of one world and the birth of a new one.

 

[DerfMidWest]

uhm. no. there are no new worlds. The game is going to get into our heads and kill us all. the world will end and there will be no other.

 

[QERB]

all i know is i use judgement hammer alot on ppl at low %s just to give it a shot, and i happen to (usually) get a lot of 9's in tournament, which can really change a game. And since the 1st judgement hammer of any stock can't be a 1 or 2, a 1/7 chance is not too shabby.

Also, getting an 8 (the freeze) can be a set up to another judgement hammer (against heavier chars), so that boosts the getting-a-nine % a lil bit in my book =P This is also helped by the fact that the 2nd hammer on any given stock can't be a 2.

 

[DerfMidWest]

Or you could be like me and be so awsome that you only get 9s trololololol

does a 1 hammer jab reset?

 

[Mr.Lemon]

Thats a good question about the 1 hammer jab reseting ppl. The only problem I can imagine would be that there may be too much lag after the side b to even do anything.

 

[Mr.Lemon]

Combo vids don't count. They only show the good bits so the proportion of nines goes way up.

Obviously. What I mean is I'll keep an eye out for any similarities for when 9's come out.

And if something was going to be the death of me I would rather it be smash than anything else.

 

[QERB]

hmm the one hammer might reset but i dont think that should ever be tactic to try n use lol... tech chasing with the judgement hammer sounds not very worthwhile when you have other options like grab/f-air, especially since grabs on almost every char can lead to another judgement hammer combo if need be. Basically, no need to take that 12%.

 

[DerfMidWest]

Tech chasing with the hammer is so funny though xD

 

[Mr.Lemon]

Yeaaa but you can just jab and know that you'll get the jab reset

 

[QERB]

it is funny =P

but ya mr. lemon's right on point lol

 

[ryuu seika]

uhm. no. there are no new worlds. The game is going to get into our heads and kill us all. the world will end and there will be no other.

Smash is the new world.

Obviously. What I mean is I'll keep an eye out for any similarities for when 9's come out.

I doubt it's going to be that simple. Remember, pre-match influences effect it too.

EDIT: Here's a thought: If the RNG really isn't random, will two G&Ws hitting on the same frame always get the same number, assuming they haven't hit previously?

 

[DerfMidWest]

you could test that with wavebirds couldn't you
hook up 1 wavebird to 2 slots and have them both be g&w and just hammer a bunch.

 

[ryuu seika]

I guess you probably could.

 

[DerfMidWest]

also smash is not the new world <_<

That would kinda suck... Random cartoony nintendo characters attacking you, bombombs being thrown around, a 2d man, etc. that would suck irl. I'd go catch a Pichu though, because he's a pimp.

 

[Mr.Lemon]

EDIT: Here's a thought: If the RNG really isn't random, will two G&Ws hitting on the same frame always get the same number, assuming they haven't hit previously?

Genius! i'll have to test this out.

 

[DerfMidWest]

yea it wouldn't be hard to test, like i said... just hook up 1 wavebird to 2 slots and press side b lol

 

[ant-d]

The mêlée PRNG is 'random enough'. You don't need decaying uranium atoms to get enough randomness for a video game

example: G&W and the opponent stand still and frame 100 produces a 9.
If that G&W does a WD frame 100 will probably not produce a 9. This is because wavedashing, among other things, advances the PRNG (changes the randomness).
Similarly, if the opponent does an attack, this will also advance the PRNG, leading to a different outcome on frame 100.

Here is a sample of some real results:
https://spreadsheets.google.com/spr...npdFc1NFpqTGJnZVJRSDFPMUpINFozWWc&output=html

Two G&W's won't get the same number even if they attack on the same frame.

 

[DerfMidWest]

I shall no longer move while playing in order to always get 9 hammers.

 

[ryuu seika]

Two G&W's won't get the same number even if they attack on the same frame.

Can you explain why this is?
If they both affect eachother's RNGs as well as their own, surely they will always have the same outcome?

 

[ant-d]

Can you explain why this is?
If they both affect eachother's RNGs as well as their own, surely they will always have the same outcome?

I haven't tried to understand the PRNG in any depth.
But, for a first hammer, both G&W's will get the same number if they don't attack at the same time.
ie. frame 100 will result in the same number for either players (if they both don't hammer at the same time, and if it is the first hammer used)

 

[ryuu seika]

So you're saying the RNG gives a different value to player 2 only when player one has had the first value, even if they occur on the same frame?
This is weird.

 

[ruhtraeel]

Computer Science student here at the U of Calgary.

Computers cannot be told to do random things. They aren't capable of it.

That's why if you pull out the hammer at a certain time, reverse the game and do it again, the same number will appear over and over again.

It's not really doing anything random IIRC, it's just following a formula, for example

Player damage + opponent damage / stocks left * time left / 1000 will be the outcome of the hammer.
even works for training mode, ex. it uses another formula to sometimes take these variables, and sometimes take other ones

It's just taking the seemingly random or uncontrollable variables of a match and putting them into a formula, I think.

So yeah, PRNG.

 

[DerfMidWest]

my brain will master this... I AM THE HAMMER O.O

 

[Battlecow]

Armada can't actually pull stitches or bob-ombs by frame-knowledge, guys.

That was just a MYTH

 

[ryuu seika]

Duh, Battlecow.
That isn't to say he doesn't know how to have a significantly better chance at them than the rest of us though.

 

[Battlecow]

No, he doesn't know how to do that either

We settled this in the peach-boards equivalent of this thread a while ago.

He just pulls turnips whenever he strategically needs them and sometimes gets lucky. He did say something once about how he "sometimes feels" like he can control what he gets, but I think he took it back later.

bottom line- it would be impossible to somehow keep track of frames, what your opponent has done, what you're doing, etc. in an actual match, much less do it and still play to your max fluidity, much less do it without knowledge of what frames stitches, etc. get pulled on. Even if it wasn't, I'm pretty sure Armada admitted to not being magic in the stitchface thread.

 

[ryuu seika]

What I'm saying is that it may, with large amounts of practice, be possible to learn instinctively how to have a slightly better chance of getting good turnips than you would have were you new to Peach.
No learning of the RNG on a conscious level would be possible, that much is clear, and there is only so much instinct can do but I think that an improved chance via years of subconscious learning may be a possibility.

 

[Mr.Lemon]

What I'm saying is that it may, with large amounts of practice, be possible to learn instinctively how to have a slightly better chance of getting good turnips than you would have were you new to Peach.

Yea but everything happens so fast in melee, and if the actions of players and everything else effects the outcome of what you pull wouldn't it be true that one minute you could be ready to pull a stitch face and the next a regular turnip.

Idk I feel it would be easier to go buy the likelihood and probability of numbers to appear rather then taking HUGE amounts of time to figure out how computer generated numbers works?

 

[Battlecow]

No. Armada would have to know what had happened, exactly, on every frame of the match thus far, know the formulae for how that effected turnip-pulling, and then execute the pull frame-perfectly, in order to have any control over what he pulled. That would be impossible if you were more than a few seconds into the match, subconsciously or otherwise.