Angel wrote: ↑Wed May 08, 2024 4:50 pm
flip wrote: ↑Wed May 08, 2024 4:14 pm
Angel wrote: ↑Wed May 08, 2024 11:52 am
Highest average rating:
lovericekim-4,95
this is the one issue i have with the tabulation (getting voters to rate each individual film is the best system, overall, but it's still not perfect). when one voter gives an average rating of almost 5, really what you're partly measuring when averaging the ratings is "did lovericekim see this movie?" the movies lovericekim has not seen do not get the benefit of their generosity.
imo the fairest system would rescale each voter's ratings so their average rating was reasonably close to the overall average rating, so that films seen by the harshest critics aren't overly penalized, and films seen by the easiest graders do not overly benefit. i'd be happy to help with the math if there's any interest in introducing that kind of change.
lovericekim's ballot is a extreme case because she only rated 39 films, mostly favorites. The system is far from perfect and this is not the best way to play, but that's life!
This ballot did raise an eyebrow for me too. FWIW I don't think we can assume that loveicekim is a generous rater. They may have felt they didn't have time to do a full ballot and just went favourite spotting on the ballot paper.
Judging on their ICM profile they mostly watch Korean movies, so that could be another reason for a short ballot, they voted mostly on the subset of very high quality Korean movies that were nominated.
I do think that for a more mainstream voter, that they're unlikely to choose 6 under the old system, "deserves to be in the TSPDT top 600" to quote from memory, because in their mind the best films are already in the TSPDT top 600.
There could be some merit in requiring people to curve fit their ballots, but I think people just wouldn't want to do that.