You can take just about any conceivable trend and anecdotally "confirm" its existence based on a small enough sample size or the right anecdote. That doesn't mean it is statistically relevant.
And yet when people make decisions, they are much more influenced by anecdotal relevance than statistical soundness. Reminds me of the time when my buddy was trying to get a spot doing research as an undergrad in one of the biochem labs. More than one prof had labs where female undergrads outnumbered the males, so he said he wished he was female so he could "wiggle my ass" to get into a lab.
edit: Also, there are statistics of overall populations and statistics of subpopulations. Sometimes the smaller sample sizes (the subpopulations) are more relevant.
Can you clarify this, are you claiming that reverse discrimination for women in CS doesn't exist? I wasn't aware that there was any question about that; I understand it to be publicly stated policy at all big tech companies. Note that I'm not making a claim that women are more likely to be hired or promoted; as I said there is a lot of discrimination against women that is less visible which is exactly what reverse discrimination is trying to, well, reverse.
I wasn't claiming anything, in terms of whether it does or does not exist.
I was just pointing out that the two anecdotes you presented are merely that. Anecdotes. I see a lot of people on HN making statements like "that claim makes sense to me because [anecdote]," without regard for how representative the sample size is. This is called cherry-picking and is a great way to (unintentionally) reinforce a flawed world view.
It might be true that reverse discrimination in CS is commonly practiced, but your initial statement did nothing in the way of proving or disproving it.
You can take just about any conceivable trend and anecdotally "confirm" its existence based on a small enough sample size or the right anecdote. That doesn't mean it is statistically relevant.