Talk:Xiphos/@comment-23.239.35.101-20160313233317/@comment-98.207.5.207-20160424102012

Chiming in on the math here, with a step-by-step explanation:

Let's say there's a 0.5% chance of each part dropping per tier. This means there is a 1 in 200 chance of each part dropping. The median amount of attempts necessary to get a single part is represented by the following:

0.5 = (199/200)^n where n is the number of attempts; solving for n results in n = 138.283

Because each part is independent and the drops are tiered, if we assume a 1 in 200 chance per part and we also assume that the player moves through the tiers as they get each part, this would result in a median of 138.283 attempts per tier. Since there are three tiers, this means the median number of runs to acquire all parts is 138.283*3 which is 414.849 total runs.

Yes, this is a lot of runs, but 1 in 8 million is both incorrect and vastly misrepresents the odds of acquiring the Xiphos. The error in assuming a 1 in 8 million chance comes from multiplying the three probabilities together. The three drops are independent, even if the drops weren't tiered. Since the drops are tiered, the 1 in 8 million probability would be the chance of getting all three drops on your first attempt on every tier. If the drops weren't tiered, it would be slightly better than this, since you'd have a 1.5% chance of getting a useful drop, then a 1% chance, then a 0.5% chance (which multiply to a 1 in 1,333,333 chance).

If you'd like a more "1 in X attempts" answer, it is simply 1 in 600 attempts (which, as explained before, calculates out to a median of 414.849 total runs). This is because it's simply 1 in 200 attempts three times in a row; that is, you'd have to make 200+200+200 attempts on average to get your 1 success (i.e. acquiring all three parts).