One problem with the current implementation of free response questions on Manifold is that there is not much profit incentive for people to submit new options, even if they think they my be right. This is because the options often start off at high probabilities, when, in reality, it's unlikely that the credence the person submitting the answer has in this single option being right is that high. This is a problem because some incentive is needed to get people to submit options despite the Ṁ25 cost and because we don't want to incentivize people to submit an option just to bet NO on it, since that would encourage people submitting obviously wrong options or just gibberish, since they know they won't pay out, rather than submitting actual plausible options.

This was partially fixed already: /JosephNoonan/will-options-in-free-response-quest

However, the options only split off at lower probabilities under specific circumstances. In particular, the very first option submitted, the one for which this problem is the most severe, still starts off at exactly 50%.

This market will resolve YES if, before November, the first option submitted in a free response question that starts out empty (i.e., the author didn't add their own options when creating the question) is immediately at a probability <50% upon addition, without having to be bet down. It would also resolve YES if a free response question that is created with only one author-submitted option has that option start off at <50%, or if there is a way for whoever is adding the option to select what probability it starts out at, and they are allowed to select a value <50%. If none of these things happen by November (CT), resolves NO.