🏅 Top traders

@traders Congratulations to @48881692 for adding the correct answer in 2 d 17 hr! The Ṁ500 prize has been sent.

The next iteration of this game is now up:

Does he have any surviving descendants?

(Since we've basically got the answer might as well ask random curiosity questions)

@TheAllMemeingEye Unknown. But likely yes.

Did he paint The Elevation of the Cross?

@ookina_inu Yes.

Based on strong evidence in favor of Rubens, I think I disagree with the response given to "Has there ever been an English-language feature-length film or TV series made about him (i.e. primarily about events in his life or featuring him as a main character)?" There's a standalone documentary titled "Rubens: An Extra Large Story" produced in 2015 which arguably meets the criteria outlined in the question. Whether you agree with this might depend on your definition of a 'feature-length film' (the documentary was 59 minutes long and vaguely obscure) but a "No" response seems misleading.

@Escher The yes/no questions can also be used to parse out my definitions of terms if you think there is some degree of ambiguity remaining. I can provide more detail as to my process with handling ambiguous questions tomorrow when I have access to something easier to type on than my phone.

For now, if I can take the liberty of rephrasing your comment as "do you consider a 59-minute film feature-length," I would answer "no."

@Escher literally my entire basis for buying 100 on no for him

@TheSchwa This seems like a reasonable definition of 'feature length', but importantly it isn't a ubiquitous one (per a google search, there isn't a precise consensus on 'feature-length'). With clarity of hindsight, asking for a rephrase (I think) would have been better.

Going meta to discuss general policy for these markets: I think that any obligation-of-clarity placed on a questioner should correspond with an at-least-equal obligation on the answerer to request rephrases on unclear questions.

Actually, maybe it's less useful to think about an allocated obligation, and more about foresight on both sides. If a questioner can foresee an ambiguity, they should modify the question. And if an answerer can foresee an a ambiguity, they should request a rephrase. Ideally both sides think about this ahead of time, but in practice people will just post whatever questions, meaning that an answerer probably has to be disproportionately aware of potential miscommunications.

A way of managing this without extensive back-and-forth might be answering questions in a "yes/no but..." format on rare occasions. E.G. "No but conditional on X definition of a feature film." A drawback of answering in this way is that questioners might assume conditioning on X is important. But the condition of X might not be generally important, and only a clarifier. Delicate phrasing might handle the problem, and in any case, having questioners decide whether to update on the fact you conditioned on X in your response might not be a problematic feature of a prediction market. If you adopted this policy, maybe some people would metagame with deliberately unclear questions, in order to try to force you to condition on some definition, thereby giving them extra evidence. But you can always just ask for a rephrase, so I don't think it would be a problem.

Thanks for getting back to me on your phone, btw. Of course, arbitrarily complex policies are easy to suggest as an ad hoc 'fix' to a specific issue, and I can see why you might not consider them generally useful. I would be interested in hearing what your pre-established thoughts are on handling ambiguity.

@Escher My overall policy is to keep the amount of rephrase requests to a minimum in order to avoid placing too much of a burden on questioners to determine the most semantically sound way to ask something. Though rule 3 states that questioners don't have to wait 8 hours to rephrase, I've found that in most cases people don't rephrase until around the 8 hour mark, which leads me to believe that they typically check in when they're eligible for their next question assuming that there was an answer provided to their previous question. In that regard, I want to avoid people feeling like they've "wasted" a question and thus I try to minimize rephrase requests.

With that as background, when I do encounter something potentially ambiguous, I basically have a two-tiered approach. If my knee-jerk reaction is that a clear yes/no answer doesn't immediately come to mind (or that I wouldn't easily find a clear-cut one online), I ask for a rephrase. However, if my reaction is that I do have an answer based on my interpretation of the question but I could potentially see other ways the question could be interpreted, I will input the question into ChatGPT and Gemini (as they, in theory at least, would represent the common vernacular). If both answers agree with my answer, I provide the yes/no response, but if there is any disagreement from either (or both), I ask for a rephrase.

Given that the jig is essentially up here, I will reveal that my process for responding to the "film" question was as follows:

1. My definitions: "feature-length film" = 60+ min, "TV series" = 4+ episodes on a topic. Are there any English-language examples for Rubens?

2. Search the web. No, but there is a 59-minute documentary. Too close to the cutoff? Would people consider this feature-length?

3. Copy/paste the question into ChatGPT and Gemini. Both say no, both provide lists of media including a feature-length film section (Belgian and Dutch examples included), both exclude the documentary from the feature-length film section.

4. Thus, return to the question and answer no.

Hindsight is 20/20 of course, but given the reasonable points you've laid out here and the implicit agreement from likes on the initial comment, I think a change to my internal policy is justified. Given that I have seen some degree of metagaming based on conditional responses in markets like this by other creators, and given my own preference for restricting responses to clear-cut yes/no answers (to provide theoretically equivalent evidence in response to each question), I think my preferred solution is that I will be a bit more liberal with rephrase requests and provide brief addendums to those responses specifically (e.g. "Rephrase ["feature-length" ambiguous]" for the film question) to preserve the equivalent evidence condition for the yes/no answers.

Was he born in either 1577 or 1578?

@robert Yes.

Is he wearing a hat in his first Wikipedia article picture?

@TheAllMemeingEye Yes.

@TheSchwa he got drip

Was he born after 1500?

@ookina_inu Yes.

Did he die in what is now modern-day Italy?

@LeonLinuxlu No.

Is his birthday known?

@robert Yes.

Did he die in a different modern day country than where he was born?

@hiperson Yes.

Was he born in what is now the Netherlands/Belgium (low countries)?

@BlackCrusade No.