Smuggler Archive
Thread: Random prediction on slicing?
ChaoticPlague
Wed Jan 19, 2005 3:14 pm
#27
*gets a good seat, cracks open a "cold one" and waits waving a Yimayz banner*
Inkanissen
Wed Jan 19, 2005 3:21 pm
#28
There is no such thing as complete randomness on a computer, random numbers are created using a formula. If you make enough slices you might get a feeling for how slices tend to behave and then take advantage of your knowledge to get a higher percentage of the slices you want as opposed to just slicing away at random.
Akiram_Glockem
Wed Jan 19, 2005 3:25 pm
#29
Is it just me or do people not understand why this thread is stickied.
Message Edited by Akiram_Glockem on 01-19-2005 05:26 PM
Cloudchasing
Wed Jan 19, 2005 3:26 pm
#30
Inkanissen wrote:
There is no such thing as complete randomness on a computer, random numbers are created using a formula. If you make enough slices you might get a feeling for how slices tend to behave and then take advantage of your knowledge to get a higher percentage of the slices you want as opposed to just slicing away at random.
OH NOOOOOO!!!!!!!!!!!!!!!!!!!!!!!!!
Please, we've ALREADY covered all the possibilities of computers a RNGs. Even if the server has some master number that it pulls all slicing data from, it certainly is NOT going to apply the next number to the same smuggler when so many are slicing at the same time. Not only that. Let's say the code uses a RNG for data, ok, thats fine. However, wouldn't the code pull all random data from the same RNG? SO, it pulls a number to determine what kind of slice, then it pulls from the same place to determine the spawn point of a lair, then it pulls from the RNG to determine what kind of Loot gets dropped. I refuse to believe that the devs (in their infinite wisdom LOL) would use a different RNG for each random event in the game. There are SO MANY random events, lets PLEASE not get into RNGs and slicing being predictable in ANY WAY.
Please, Dreven, get back to us with your test. Please, I have total faith that Samaw is just really lucky, but I don't believe that it's predicatable at all.
Nicolas_Frost
Wed Jan 19, 2005 3:30 pm
#31
The percentage is not the only part of the slice that is controlled by the RNG, so is the type of stat increased. The difference here is there's only two stats on an armor piece/weapon that can change, meaning its a 50% chance for damage/effectiveness or speed/euncumberance. Essentially, a coin toss. And while you can flip a coint a hundred times and it land heads each time, its #1: Unlikely, and #2 Neither guaranteed to be half and half. Each time you pick up that coin, your chances of it being tails do not increase nor decrease, they remain the same.
Nicolas_Frost
Wed Jan 19, 2005 3:32 pm
#32
Cloudchasing wrote:
Inkanissen wrote:
There is no such thing as complete randomness on a computer, random numbers are created using a formula. If you make enough slices you might get a feeling for how slices tend to behave and then take advantage of your knowledge to get a higher percentage of the slices you want as opposed to just slicing away at random.OH NOOOOOO!!!!!!!!!!!!!!!!!!!!!!!!!Please, we've ALREADY covered all the possibilities of computers a RNGs. Even if the server has some master number that it pulls all slicing data from, it certainly is NOT going to apply the next number to the same smuggler when so many are slicing at the same time. Not only that. Let's say the code uses a RNG for data, ok, thats fine. However, wouldn't the code pull all random data from the same RNG? SO, it pulls a number to determine what kind of slice, then it pulls from the same place to determine the spawn point of a lair, then it pulls from the RNG to determine what kind of Loot gets dropped. I refuse to believe that the devs (in their infinite wisdom LOL) would use a different RNG for each random event in the game. There are SO MANY random events, lets PLEASE not get into RNGs and slicing being predictable in ANY WAY.Please, Dreven, get back to us with your test. Please, I have total faith that Samaw is just really lucky, but I don't believe that it's predicatable at all.
Exactly. I just used a small portion of what's available in my demonstration.
Cloudchasing
Wed Jan 19, 2005 3:35 pm
#33
I was just listing some of the randomness. SUre it's 50/50 for what kind of slice, PLUS it's 50/50 for what kind of wire gets cut.However, how many other random things are happening during the time of a slice. Each slice calls for3 different numbers. Whats percentage I'm sure has some long (well, not all that long really) equation based on a number pulled, but that's3 different pulls from the RNG. Adn those are going up against each random event in the game. So predictability is practically impossible.
We'll hear from Dreven eventually.
Heck, I have a few crates of knives in my house as well, Maybe I'll try this out too.
riotcontrol
Wed Jan 19, 2005 3:37 pm
#34
TomoRainer
Wed Jan 19, 2005 3:40 pm
#35
10-12 slices isn't any kind of significant sample. If you got 80% predictive power on 10-12,000 slices, or even 55%, now that would be something. But 10-12 slices? That's like saying "Well, Mike Cameron just hit 5 home runs in his last two games. Yep, looks like he'll end up with 900 or so by the end of the season." Sample size is critical to this kind of thing, and 10 is so untrustworthy as to be literally useless. If Dreven slices 100 items, that'll give us a better idea, but still not an entirely reliable one. Our stance that slicing is random comes from both developer word and from experience: multiple smugglers have collected thousands upon thousands of slices, and the collected data show an extremely random distribution. If you come up with a method to challenge that, fine, but at least put in the effort to collect a significant sample before waving around the latest voodoo theory.
Samaw
Wed Jan 19, 2005 3:50 pm
#36
Let try to explain what I have noticed a different way. Because in no way do I claim to have figured out the RGN or what the seed is.
What I have noticed is that when slicing large amounts of weapons that the same color gives the same kind of slice and the opposite color gives the other slice for about 20 to 30 seconds and then changes again. Only if you slice the same kind of weapon. But if you slice two different kinds of weapons it tends to be the opposite again this is not full proof, but it greatly increases the chances.
I know this topic has been beaten to death, but I have seen this work for me and just want to pass it on to you all. So if you wantto continue tonotgive it a try then don't. But if you want that extra edge then do a little bit of testing for yourself.
Message Edited by Samaw on 01-19-2005 05:58 PM
Zhagadska
Wed Jan 19, 2005 5:15 pm
#37
Samaw wrote:
Does anyone else feel that they can predict the weapon slice with about 80% accuracy?
The other day using tell slices I was able to get 7 out of 9 damage slices with some Acklay Power Hammers.
Message Edited by Samaw on 01-19-2005 03:21 PM
oh no you di-int
MISTagent01
Wed Jan 19, 2005 5:16 pm
#38
Plain and simple, if you can't predict it with 99.99% or 100% accuracy, it's not a pattern.
TomoRainer
Wed Jan 19, 2005 5:26 pm
#39
To be fair, by definition any slices you could predict more than 50% of the time would be a pattern. Wouldn't be very useful unless you could do it at least, say, 60% of the time, though.
And for the record, I don't think any patterns will ever be found.
And for the record, I don't think any patterns will ever be found.