Greetings fellow gamers! A friend that is an avid D3 player asked me to render an opinion on some threads dealing with RNG here, which led me eventually to this thread. I thought I might be able to contribute something useful, so after pinging the thread creator re: linking to elsewhere, I humbly offer the following.
My education and background of several start-ups had me intimately involved in the subject at hand. Seeing the same kind of misunderstandings I see here in the fora I frequented, I decided to do something about it. I produced some guides with the intent of giving the interested reader something to chew on beyond the typical post, beyond wikipedia, but less than a textbook. I tired to condense what I would teach in a semester into a few dozen pages, and focused on how it relates to games. I hope at least a few here find it interesting and useful.
There are two guides in particular relevant to this thread. Neither is short. Both require some effort and willingness to think on the part of the reader. I am convinced, however, that the reward of that effort will be an understanding for the reader well beyond that of the typical poster. There really isn't a TL;DR version - these *are* TL;DR for getting a degree in CS/Mathematics. If you aren't really interested in the facts and putting on the thinking cap, they probably aren't going to be your cup of tea. With that in mind....
A Guide to RNG and its use in games- This covers what an RNG is, what most games use, how this is combined with game logic to produce drops, and why there's "good" and "bad" luck, streaks, dry-spells, etc. in random systems. There are code examples (no worry if you're not a coder - the results are explained).
A Guide to Proper Drop Rate Inference and Sample Sizes- This covers the mathematics behind making proper conclusions about drop behavior from observations of the game. This is a bit more mathematically slanted than the RNG guide by necessity, but nothing anyone comfortable with algebra would have trouble with. This addresses the all too common fallacy of small-sample comparisons and unjustified assertions from small samples.
Happy to answer any questions via whatever mechanisms are appropriate here (PM, thread interaction).
0
My education and background of several start-ups had me intimately involved in the subject at hand. Seeing the same kind of misunderstandings I see here in the fora I frequented, I decided to do something about it. I produced some guides with the intent of giving the interested reader something to chew on beyond the typical post, beyond wikipedia, but less than a textbook. I tired to condense what I would teach in a semester into a few dozen pages, and focused on how it relates to games. I hope at least a few here find it interesting and useful.
There are two guides in particular relevant to this thread. Neither is short. Both require some effort and willingness to think on the part of the reader. I am convinced, however, that the reward of that effort will be an understanding for the reader well beyond that of the typical poster. There really isn't a TL;DR version - these *are* TL;DR for getting a degree in CS/Mathematics. If you aren't really interested in the facts and putting on the thinking cap, they probably aren't going to be your cup of tea. With that in mind....
A Guide to RNG and its use in games- This covers what an RNG is, what most games use, how this is combined with game logic to produce drops, and why there's "good" and "bad" luck, streaks, dry-spells, etc. in random systems. There are code examples (no worry if you're not a coder - the results are explained).
A Guide to Proper Drop Rate Inference and Sample Sizes- This covers the mathematics behind making proper conclusions about drop behavior from observations of the game. This is a bit more mathematically slanted than the RNG guide by necessity, but nothing anyone comfortable with algebra would have trouble with. This addresses the all too common fallacy of small-sample comparisons and unjustified assertions from small samples.
Happy to answer any questions via whatever mechanisms are appropriate here (PM, thread interaction).