[wongarsu]

>Likewise, if all you know about someone is they are a developer, assuming they are male is not a terrible strategy since you are correct 90% of the time!

Yes, and I'm not questioning that. The problem is that that same statistic says nothing about the likelihood that a given female is a developer.

[alexmat]

But if about half the people in the world are female and there are more male than female developers, then isn't it safe to assume that the likelihood of a random male being a developer is going to be higher than that of a random female?

[wongarsu]

When picking a random person from a phone book, yes. But that's not what this whole discussion is about. In virtually any real situation you don't have a completely random female. In most relevant situations you don't even have even close to equal amounts of men and woman in your sample.