Armorsmith Archive
Thread: To AS on hiring slicers and the random # generator in this game
Page 1 of 1
Istin
Thu May 20, 2004 7:39 pm
#1
Hey all, almost sorry to intrude on your forum, but the smuggler one is feeling a little retarded atm.
Anyhow, here's some interesting information I picked up/realizedin the days of my slicing and it ALL branches off of the random # generator implemented in this game.
This generator has STREAKS, for SURE. If any of you are pistoleer this is noticeable in short spurts when fighting quenkers or something, it's more or less a series of dodges, followed by a series of hits (on you). As a slicer/Msmuggler, I noticed longer streaks in general, as I sliced a great armorsmith's wares on Shadowfire.
Basically, if your slicer is getting rather bad slices (if they care, you should ask them to keep you updated on how it's going) tell them to stop for a minute or two, or a while. I can gurantee that if they are doing "bad" they will keep doing bad.
I was getting a series of bad slices for an entiremorning once.A bit later (still near noon) I have to start slicing for my AS employer, and he's working on getting a 80% kin, 90% base set. Early in the first crate,the slices were definately not going so well (about 10 pieces in) so I tell him I'm not doing so well, maybe another slicerwill have better "luck" atm, he says keep going. I do 5 more pieces, still not nearing the quality I like. I think I waited about 15-20 minutes, slicing a piece from that crate every now and then. FINALLY I hit the magical 31% effect slice to make the piece hit 90% base. Almost all the slices later (about 225 more) are made with "good luck".
At the end, he gives me a 8 crate of stun helms, kinetic chests, and 3 sweet krayt helms (which all of you know end up being all special, needing encumb slices). So I start in on the chests, not doing too great, but what the heck, it's almost over, I'm just looking for and enumb slice pattern. Nothing through the chests, start in on the stun helms. At last, I get I think 4 or 5 encumbs in a row. So, of course, I hit the krayt helms. first slice.....effect 
. Second slice 43% ENCUMB. Figurnig I've used up my "luck n encumb, I finish off the stun helms and last krayt, which, btw, got a 33% encumb slice.
. Second slice 43% ENCUMB. Figurnig I've used up my "luck n encumb, I finish off the stun helms and last krayt, which, btw, got a 33% encumb slice.Anyhow, those stories were merely to give an example of this generatorat work.
Point is...have your slicer make a crate of survival knives and have them slice about 5 every now and then. Then, when they hita fewdecent/good slices in a row, have them come over and start work. Sure, it's a little inconvenient, but it's worth it when you can sell your better sliced armor at higher prices.
Also, the same goes for your experimentation. If yo're getting a high rate of amazing successes, start crafting like crazy!!! Might only make a few points of HAM lower or a base effect 1% higher, but when you start having the slightly better armor, always, you can raise your prices exponentially (or so I've discovered).
I tried posting these results on the smuggler forums, but it doesn't seem they care to much, maybe it will to you Armorsmiths.The smugglers arealso working on our revamp doc, so we're all a little distracted atm too.
MrMud
Fri May 21, 2004 5:14 am
#2
Before anyone takes the time to do a full statistical analysis of the phenomenon i would just blame it on lack of intuitive understanding of statistics
Istin
Fri May 21, 2004 8:32 am
#3
No s**edit**t Sherlock. Of course you can't have a control in this game, on anything involving randomness. Everyone should knowthat.
I wasn't trying fora "full statistical analysis. I was justTRYING to be nice and tell y'all that:
a) The random # generator in this game really isn't truly random, it's far too streaky to be random. Or would you call 2hrs of good slicing, almost all slices above 20% effect/30% encumb random? I mean, that's almost 40% of the etire slicing range barely happening over a set of225 slices.
b) If you want the best results on anything you craft, watch carefully on experimentation or slicing. Stop for a while if you're doing bad, go as fast as possible if you're coming out with really good results at the time. Don't stop just 'cause you're done with a particular project, go craft some layers or segment schematicsthat might be a little better than the ones you already have.
It was a long post 'cause of the stories, of which I haveMANY similar examples. I thought they might lend a little credence to the theory.
Don't flame someone you don't know ANYTHING about, go out and try it, they might be right. You should have been able to tell by my spelling, punctuation, use of paragraphs, etc... that I at leasthad some common sense.
ctrl-F1
/flameblanket
MrMud
Fri May 21, 2004 11:42 am
#4
I dont think you understood what i was trying to say.
First of all i was not trying to flame in any way. It is just my experience that humans are very bad at judging statistical phenomenon (this includes me).
My point was that you do remember the streaks and you most clearly notice the truly bad and good streaks not what usually happens. And most of the time when you look at this from a statistical perspective it makes complete sense with what a reasonable mathematical model would predict.
FreeEnterprise
Fri May 21, 2004 1:21 pm
#5
It is impossible to generate truly random anything with a computer.
BaseballMartinez
Fri May 21, 2004 2:45 pm
#6
But it is possible to do statistically random numberswith a computer.
DHorse
Fri May 21, 2004 3:40 pm
#7
I totally agree with Istin. Eventhough I am fairly new to master of Armorsmith, but the other day, when I built 5 set of ubese for my guild, same thing happen. I was building the layer (kinetic.) Couldn't go pass 10% on special at all, after 5 attemp, I got mad and decided to go hunting. After couple hours, I decided to go at it again, and I got 14% on the layer, this is with all the same ingridients. I have tried the same idea when I build the segment and glove, once I got the idea, I built all the schematic for the Ubese suits. Now I can have 1000 of the 74% kinetic Ubese.
BTW I am on shadowfire as well.
JaatoWaals
Fri May 21, 2004 3:53 pm
#8
What MrMud is referring to is called an availability heuristic. It's one of the many reasons why humans are notoriously bad at statistics especially when they are based on anecdotal references. I'm inclined to agree with him- without some kind of formal statistical analysis you're going to have a hard time persuading anyone with an understanding of how randomness works that the system isn't actually random, given a presumption of randomness (i.e.: the devs told us so).
Istin
Sun May 23, 2004 9:05 pm
#9
Alright, sorry I jumped on ya' MrMud . Guess this game's forums made me a cranky old man 
. Guess this game's forums made me a cranky old man Ok, yeah, I KNOW I can't do a formal statistical analysis, I don't have the time, drive, or knowledge/recall for that. HOWEVER, this post was meant to be a suggestion for everyone to try what I'm saying.
I'm saying this because I've sliced easily over 2000 pieces of armor or weapons since I became Msmuggler. Of course, I figured I'd try to be "the best" in any way possible, and the general consensus on the smugler forums is that everything is random in slicing and no one can be better than another. Well, hearing that from the day I mastered, it's my human nature to try to beat it/disprove it. So I started keeping a mental track of what the variables might be and what effect they might have, starting with what has been said in the smuggler forums.
Check out the many pages on the smuggler forums and you'll see countless posts about "has slicing changed?", or "did they reduce slicing values?" Followed by posts abot how this one person has been having an amazing streak of slices and that maybe the devs did a "stealth...."nerf"" (for a lack of a better word), or how someone's been having consistent bad slices.
After all those stories and my own first 1000 or so slices, I started to wonder, and even attempt to keep track mentally of the general trend of my slices. The....anecdotes....in my first post were mere examples, albeit some of my stronger ones, of this trend I'm noticing.
I guess what I'm REALLY trying to say is....It may be too much trouble to have your slicer "test" his slices at the moment and all. You may not care. But just TRY it, experiment a little. If your slicer is having "bad luck", have him keep slicing and start telling you the slices, you may be surprised.
Don't try it if you don't want, but I know there will be SOME hardcore crafters out there that strive for the best equipment of any type, and if they try this trend and it improves their wares by even a little, I'm sure they'll be happy. Sure, many slices will be average, but after slicing hundreds of pieces, and the overall quality is improved, they and they're customers will be happier.
Page 1 of 1