User blog comment:Dragonjet/KanColle English DB/@comment-24835439-20140419151825/@comment-24816414-20140419170233

Well, if you what to give an approximation of the drop chances, "No drop" and "didn't notice" should be considered in the calculation.

Using your example: if they aren't used in the calculation Iku's rate would be 100%, 1 sortie -> 1 drop (100% drop rate). But if you use them she would have her real % value, 10 sorties -> 1 drop (10% drop rate).

And the fact that other ship drops makes a drop of a certain ship go lower is also a problem. So you should consider the other drops too...

umm, what about this: to estimate the drop rate of a ship you should take in consideration: number of sorties, no drops, drops that aren't reported, other ship's drops and the drop of the ship you are interested. So maybe you should use something like: % drop of X = reports of X /(No drops + didn't notice + reports of no X). Using this way you also  take in consideration all those people who didn't notice their drops or were lazy to registrer common ships or Nakas. This way we get a somehow good aproximation I think.

Then there wouldn't really be any diference between "No drops" and "I didn't notice" so you could just delete one of the options.

I am saying all this but, well, I am just trying to help. You are the one doing all the work so if this is hard or a pain you can just ignore it.