[cooldoganytime]

I love all forms of media (films, series, games, books, music). I enjoy challenging my taste and discovering new things. My issue with popular media recommendation platforms is that they tend to be watered down with mainstream titles (especially those available on popular streaming services), they make the same obvious recommendations over and over, and they shy away from old, weird, and obscure titles. It doesn’t help that these platforms are largely owned by massive media corporations (Amazon, etc.) which don’t have a vested interest in introducing you to something unique and interesting (in fact they have a vested interest in corralling you towards their own content).

I also love to read discussions that allow me to engage with media on a deeper level. I like to see different perspectives/interpretations or learn about details that I missed. I appreciate social networks like Reddit where some discussion takes place, but in-depth discussion has been largely replaced with low-effort and tired memes and catchphrases.

My goal with this web app is to create a niche community of open-minded individuals who love to discover, share, and discuss media outside of the mainstream. The community is self-governed and self-sustained, with no reliance on external APIs or services, and with an emphasis on quality over quantity in both recommendations and discussion. It is completely free with no ads or data-harvesting. The app is feature-complete and I officially launched it today.

I’m a self-taught hobbyist, and I built this app over a few years with Django and Vue. If anyone has any feedback or questions about development then I’d love to hear them.

[purplecats]

How does recommendation work? There used to be a site called mov3y which was also on HN a while ago, and I loved it. this gives me similar vibes. that site did recommendations by finding users who also rated movies you have rated about the same score and i found that to work really well.

[cooldoganytime]

I haven't heard of mov3y, but it sounds very similar. On peepthis a 'compatibility score' is routinely generated and updated between you and other users, as determined by the similarity of your ratings. Your 'predicted ratings' for titles are also routinely generated and updated, as determined by the existing ratings of your most compatible users. So different users will have very different recommendations based on what users with similar taste like, rather than having monolithic "best of" recommendations that are only based on average rating and popularity.

I was heavily inspired by www.criticker.com which is the best site for film recommendations that I've found and I've been using it for over a decade. However it's feeling pretty dated at this point and I also wanted to apply the same concept to books, music, etc. which is why I decided to create my own platform.