Quote:
Originally Posted by L33TNeMiSiS
Strcpy: I'm one of those guys as you can see from my post. If you like. I'm willing to meet you in GW, do any run and show you how much your/other ppl drops will differ from what I get with my Character. Maybe we should make it a bit of an event :P. Get all the guys that say they get bad drops and then get the guys that say they get the 1 gold/hour and record everything. Maybe settle this debate that way
|
If there was such a thing as a Law of Averages than your idea would be somewhat useful, except there is no such thing. The closest you get is the Law of Big Numbers and that is what I am trying to achieve with the above protocol. Even should we do it and you get all the drops (me being "correct" as I never play those areas as I am a pure H/H or solo player) it *still* wouldn't mean anything.
Every single time someone has done the above they get *no* anti-farming code that you complain of and that is irrespective of how much they believed in it. At one time I kept track of my Bergin Hot Springs runs "prove" that there was one in effect and found out I was incorrect. This rumor has been around since day one and there has yet to be anyone find evidence of it once they start applying a systematic approach and Anet has denied having this in there since then also.
Quote:
Originally Posted by Musei Karasu
I believe the issue is not that humans make patterns out of the random, but instead that a computer cannot create a "random" no matter how much we try to do so. There is eventually a pattern. And just like you said, humans are extremely skilled at finding patterns. We see that pattern and think it is something that has been programmed when in reality it is actually the limitations of computers.
|
What this means is that you can not predict that next state generated from looking at the previous ones (and thus, for all intents random) but if you knew the algorithm used and the seed you could predict each successive exactly. This doesn't cause a single individual to experience anti-farming code in any way and has nothing to do with it. Even with the *REALLY* poor ones that you are going to find a pattern by looking at the outcomes they take large amounts of data and generally complex statistical analysis to find the pattern (and since PRNG's are central to much of gaming they do not use the bad ones and haven't since the early Atari 2600 days). And, even then, all it would allow you to do is to predict what is going to drop next, not have a systematic depression of drop quality against a single individual.
You will note that a fairly famous one (http://en.wikipedia.org/wiki/RANDU) takes thousands of samples to notice and requires the values to be plotted in a specific way. It is considered to be one of the worst ones out there and you could not detect that pattern just by watching numbers come out (or drops generated from it).
Even should they seed off of "time spent in game" it *still* wouldn't degrade drops for long term players unless they used a *really* poor PRNG and, again, the field is old enough that it is well known how to avoid this - it takes intentionally doing so even if there is one out there (I know of none as that destroys the whole statistically random thing). Plus it puts us back into the Anet is liars and it violating any of the longer term data anyone has compiled.
All of this is drilled into your head on any course dealing with seminumerical algorithms (go have a read of Knuth's volume two of The Art of Computer Programming/Seminumerical Algorithms). Your taking a small tidbit of information and carrying too far - you know enough to extrapolate some *really* incorrect ideas.