Another FB engineer here - I actually work on the system that caused this false positive. You're right that if we were actually trying to stop constructive dis

Another FB engineer here - I actually work on the system that caused this false positive. You're right that if we were actually trying to stop constructive discussion from happening, that would be bad. We're definitely not trying to do that - in fact, our goal is the exact opposite. Similarly, if we let through blatantly malicious or spammy comments, I think that would be bad as well. While I wish I could say we were perfect at stopping spam, the reality is that no spam classification system is perfect. Sometimes we err too far on the conservative side, and spam gets through. Sometimes we err too far on the aggressive side, and good content is incorrectly blocked. That's what happened here - it turned out one of our spam classifiers was a little too aggressive, and we've turned it off.