We do filter duplicates, but only when the story has had significant attention in the last year or so. This is in the FAQ: https://news.ycombinator.com/newsfaq.html#reposts.

That leaves two cases where the dupe detector is deliberately left porous [1]. One is if the story hasn't had significant attention yet. We want good stories to get multiple chances.

The other case is when a story has had significant attention, but not in the last year or so. In that case we want to allow reposts because some stories are of perennial interest and a year (or so) is enough time for the hivemind cache to clear. Also, it's good for newer cohorts of users to get exposed to the perennials and classics. That's part of how the culture propagates.

[1] https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...