I think that you should close voting two weeks early. Heck, make it three.

[Fredrik]

Eric Morton wrote:

Vic Wertz wrote:

Second, a "reject" button would actually be a harmful addition from a data-gathering point-of-view. This system works best when entries are paired up against the widest possible variety of other entries, and if you could opt out of seeing a particular entry, that would result in fewer pairings being registered.

That's why I'd like to see a "Tied for Last Button" (with a large but finite number of allowed uses per voter). It would automatically provide data for every possible pairing including the "tied for last" item, while simultaneously guaranteeing I never have to see it again.

That would provide data in the sense of filling fields in a database, but not in the sense of making an analysis of the data more useful.

[Eric Morton]

Fredrik wrote:

That would provide data in the sense of filling fields in a database, but not in the sense of making an analysis of the data more useful.

Assuming the "Tied for Last" button had a limited number of uses per voter, it would open up all sorts of interesting, new ways to analyze voting trends. Sure, you'd end up get a bottom tier of items that weren't ranked relative to one another, but a similar loss a data happens when you drop the bottom tier of items from voting less than half-way through the voting period (which has already happened this year).