In a previous guest post I had argued that we should use a random-sample-voting algorithm in any kind of system that promotes certain types of content (songs, tutorials, ideas, etc.) above others. By tabulating the votes of a random sample of the user base, this would reward the content that objectively has the most merit (in the average opinion of the user population), instead of rewarding the content whose creators spent the most time promoting it, or who figured out how to game the system, or who happened to get lucky if an initial “critical mass” of users happened to like the content all at the same time. (The original post describes why these weaknesses exist in other systems, and how the random-sample-voting system takes care of them.)
However, this system works less well in evaluating the merits of a rigorous argument, because an argument can be appealing (gathering a high percentage of up-votes in the random-sample-voting system) and still contain a fatal flaw. So I propose a modified system that would work better for evaluating arguments, by adding a “rebuttal takedown” feature.
Continue reading ‘A Modified Algorithm for Evaluating Logical Arguments’