Suggestion: Heavier weightings on higher films

[jwaycuilis]

I think that in determining a TCI, movies that you have the highest rankings for (which you are probably passionate about) should have a higher weighting. For example, if someone shares absolute favorites with me, I would like them to be one of the highest rated people on my TCI list.

[Quicky]

This suggestion has come up in the past and I still support it. In science you should always give more weight to data that is more accurate. In this case I believe that the highest tiers as well as the lower tiers are films you feel strongly about and will thus be more accurately rated than the hordes of so-so movies that you don't really care about. More accuracy = more weight, less accuracy = less weight. Simple.

In a similar fashion, TCI's that have rated a lower amount of movies should also be given smaller weights since the accuracy of a TCI correlates with the amount of movies rated. A TCI of 1.6 with 50 movies rated will be less reliable than a TCI of 1.8 with 500 movies rated.

A weighting like this for the tiers could be optimal:

Code: Select all

10: 2.00
 9 : 1.40
 8 : 1.00
 7 : 0.85
 6 : 0.70
 5 : 0.70
 4 : 0.70
 3 : 0.70
 2 : 0.85
 1 : 1.00

[mpowell]

This is a suggestion that often rears its head, and one that is under investigation. We've resisted it so far, out of a desire to keep the formula for TCIs simple, but we are planning on looking into it.