[new] add new method of TCI/PSI calculation respecting nominal user ratings

[livelove]

As explained here and here, I am unsure if comparing percentiles to each other is really the best idea — for the reasons outlied in these links, especially the fact that users try to avoid seeing bad movies, which, by the way, is the whole purpose of this website.

As users don't like to watch bad movies, the current situation leads to absurd aberrations, like users feeling obliged to badly rate unseen movies or copy bad ratings from other users.

suggestion: You could do both — compare percentiles AND actual votes against each other.

Each method yields TCIs and PSIs.
You can then compare the prediction accuracy of both methods and see which one is more accurate!
So there is no need to guess. You could let science sort it out itself.

I would be dying to see this.
This is actually THE one question I have wondered about ever since I became a member here.

suggestion: You could let the users choose (per setting in their profile page) between seeing TCIs/PSIs based upon percentile comparisons or actual votes comparisons OR BOTH.