Another new bit is now for unrated (via new system) titles you see a percent match instead of the average rating of other subscribers. I have not got a feel for how accurate this is yet but again that would infer they are trying get the model even more accurate. Generally the way to do that is to get finer detail of the watcher's like and dislikes not less.
The other issue with a binary system of course is what about the stuff that is just OK? As in stuff that was worth watching once but not again. How do I rate that with love or hate as my only options? Marking it either way will skew their suggestion model and not marking leaves me no way to see I've already watched it. To look at this I downloaded my "My Activities" page with over 12 years of ratings and converted it to a sheet so I'd have the data. And to see how is much this might truly effect things. I found:
- The single list of ratings on "My Activity" has entries formatted 3 ways which made converting it to a sheet a bit of a pain. (This might also mean a previous change in December of 2015 is affecting this.)
- Of the 2160 titles I've rated 1897 (~88%) were in the 2-4 star range.
- 947 (44%) of my ratings are 3s that the new binary system can not handle.
- 138 of the movies and 23 series I have rated are no longer on Netflix.
- My average rating under the old system is 3.2
- Converting thumbs down to 1 and up to 5 my average rating drops to 2.5 with the new system
- So on average titles lose about 0.7 stars under the new system
Facebook removed their rating system for movies and TV shows awhile back too. While you can still add them to watched / want to watch sections and even write a review as a comment, all those star ratings you did are now toast. Funny thing is when you try to find out when did this there appears to be no stories about. Probably why Netflix thought it was a good idea to follow suit. Unfotunately for Netflix people are noticing.
No comments:
Post a Comment