Quote: Originally Posted By colonels19 on 10/24/2009
Quote: Originally Posted By gin_caesar on 10/24/2009
"oddly cluster for certain periods of time just because" It happens, and other users have said so in this very thread.
I gave you too much credit. I assumed that you understood what was going on in your own example which you paste all over the place. My initial instict to just call you a moron appears to have been correct. Gin, we agree that there are implementation of randomness problems in this sim as well as other sims, I'm just taking the argument a step further because I don't think TRUE RANDOMNESS is used, which would further compound the problem. Again, I don't see how/why this is so hard to understand...its very possible for WIS to be using SOME SORT of PSEUDO RANDOMNESS, especially with the clustering problems if/when there are 16 MILLION potential numbers/outcomes.
When I told you I was an air traffic controller you were soliciting advice on a malfunctioning mouse or some other **** that for whatever reason you thought I'd like to help you on.
To be honest, I'd rather be told to f**k off than to be lied to, in any situation. A long time back, 3-6 years or so, I remember that you told me in a post somewhere that if I needed programming help, that I could ask you, thus that's what I was doing when I sitemailed you. I don't hold grudges, I don't hate anyone on this site, and I'm just as open and willing as anyone else on this site to help someone else if help is needed and asked for. Again, I don't consider "thread hate" to be serious or to even matter, that's why I was looking to you for help, and that's the honest truth. And FWIW, it wasn't a mouse question, it was something to do with Microsoft Access I believe, but I can't find the original sitemail. Take care, gin.
I'm going to reset this and be as non-technical as possible. First let me tell you that I'm pretty sure that they would use pseudo random numbers, but don't freak out...
There isn't a concrete pseudo random number generator. It's an umbrella term for a method/twister/anything that makes random values from something that isn't random. There's a watchmaker discussion to be had here but I'll save that for another time. But you see, implementation is all that really matters with pseudo random numbers because onus is all on the developer to ensure that there aren't problems. As long as collisions/repeats/<insert faults here> are handled at the code level there is absolutely nothing wrong with pseudo-random numbers. There isn't an inherent fault that is definitely going to happen like "clustering" because there isn't a standardized concept to be followed when generating the number. I say this first as somebody who in ten years of development has never worked hands-on with true random numbers and never needed to even in things far more complex than a basketball simulator, and second as somebody who thinks that if a developer isn't accounting for how a prng is affecting output, true random numbers aren't going to save him because he doesn't know what the hell he's doing in the first place.
See the link you posted, as I cleared up earlier the developer didn't use the best method implementation and that was the cause of the repetition, not the method itself.
Make sense any of this?