layering prediction markets on trading outcomes attract a disproportionate amount of attention as it is much closer to the heart (or rather the pocket) for most people e.g. finding out a particular model is better at coding or writing research paper, a layman's reaction will be 'aight cool'. most of the time they don't actually understand the benchmark in the case of @the_nof1, it's simple: who is making the most money? the outcome itself is not only simple, but also highly speculative (the nature of trading) it also gives non-traders the feeling of 'oh wow even without being a trader i get to speculate on trade outcomes'. making accurate guesses on which model will be able to perform the best becomes a 'vanity metric' for people: 'yo look i guessed this correct' as i like to put it: multiple choice questions instead of open ended