First, randomly drawn an affix from 178 affix as first attribute, the affix and delete that class library, continue to drawn from a second affix the rest of the affix, until all five properties are determined.
Second, affix a random sample of five classes, randomly selected affix one from each class as the final properties.
Buy d3 gold think if blizzard programmers is human, then take the second possibility is bigger, no matter from the code design and implementation are much easier than the former, so the original poster tendency as the second possibility. But the original poster is not sure blizzard programmers are human, so need to decide it.
Most of the above affix no matter what kind of situation is rarely seen in small sample distribution difference, so choose a special class with hole here.
If for the first, the article 500 of affix in the hole are expectations: 500*3/178≈8.427
Which is 100 about eight to nine necklace necklace has a hole.
If for a second, then 500 affix in
buy diablo 3 gold hole expectations for: 100*C(26,4)/C(27,5)≈18.52, that is about 18 to 19 in 100 necklace necklace has a hole. As you can see this is very obvious differences, we can through a small simple samples to determine which hypothesis is more likely for the actual situation.
没有评论:
发表评论