Board Thread:Question and Answer/@comment-24125853-20160418134247/@comment-73.3.99.102-20160423021752

I'm just speaking hypothetically here. Suppose you have a drop rate of 2% or 0.02 for any AW, I'm just shooting from the hip as well as it's been a while since I've thought about combinatorics and independent vs dependent statistical events.

However, and getting back to my original hypothesis. If we assume (generously I might add) a 2% drop rate which is the equivalent of 1 drop every 50 possible drops. And we also assume that both Drops from an F/AW kill are possible drop events then the result of 2 drops occurring simultaneously for any F/AW kill is 0.02x0.02 or 0.0004 or 0.04% roughly 1:2500. Now if we assume that 1 AW drop negates the possibility of another drop from an F/AW kill then that probability will automatically compress to 0 due to he second event being dependent upon the first.

That probability also goes down if we look at the samples above that were provided if we assume a drop rate of even 0.01 then the likelihood of 2 FAW drops in a single event is 1:10000, meaning you would need a huge sampling of kills to achieve a double FAW drop. I don't know of many players that kill 5k F/AW's.

Now, let's assume for a moment that some random player gets so lucky they actually recieve one of these drops. Maybe 1 out of the 7,000,000 players gets one. What are the odds that player posts their luck to the web? Or is reading and posting to this chat?

My point is this, not seeing something and something not having a possibility are two separate things. I've never seen sea turtles or unicorns. What are the odds? I'm being a bit facetious, but still...

I am wondering if we new the number of witches, as well as the probabilities for drop of each type, we could get a far better permutorial representation of the probabilities based on a choose function. For example if there were 5000 total cards and two cards had a probability for drop. Then we could use a similar function to what is used to calculate the probability of a royal flush in poker.

There may be system locks against such events but so far I've not seen an argument compelling enough to make me believe that both drops include a possible F/AW drop.

One other thing that just occurred to me suppose the drops were not independent events and the first drop did not result in an AW drop. Then the probability for an AW drop would be double or 0.04 for each F/AW kill. However the probability of a double drop would be 0 as the drops depend on whether or not an AW dropped in the first drop.