Training a spam filter is not the same as creating a do-not-deliver rule for a given sender. Unless you manage to convince the filter that paypal is the same as

Training a spam filter is not the same as creating a do-not-deliver rule for a given sender. Unless you manage to convince the filter that paypal is the same as a viagra scammer (good luck with that), you'd still get those messages from paypal. Maybe some providers also denylist the sender-receiver pair when you click spam to combat this problem nowadays; back when I used public email services this was rarely the case. Nowadays I use a different system altogether so I don't know if this might now be common. Either way, this is not what the spam button is for, but from this subthread I see that enough people on HN already don't understand how these systems work (and education is hard: people don't even read relevant oneliners pushed in their face at a relevant time, such as error messages), so I guess there's no hope for the general public altogether. I didn't know this is an entirely lost cause and is making me rethink about reporting spam on platforms like reddit. I guess they get so many false positives that I'm wasting my time reporting anything as spam ever.