Scott Alexander

This was technically framed as a recommendation to Oracle voters, but it is so effective in establishing the shelling point that it's practically always followed.

However, in this case, there were only two minutes left, which wasn't enough time for the voters to change their mind.

Seeing that the resolution was trending towards yes, the Polymarket representatives, not wanting to break their streak of always establishing the shelling point, changed their own opinion to yes, and the final vote was yes 99%.

Doma, how many people watched the Oscars on 3-5-25?

Kalshi's resolution criteria for this market said that they would resolve it when a major news source published Oscar viewership numbers.

A few minutes after the Oscars, New York Times published preliminary viewership numbers without any caveats saying that they were preliminary.

The next day, they published another article saying that actually, the real viewership numbers were higher.

Kalshi decided that the letter of the resolution criteria was met when the New York Times published its first article.

There was nothing in the resolution criteria saying that they would be corrected if a new source changed their mind.

Therefore, they would pay out according to New York Times' first preliminary, that is wrong, estimate.

Traders who bet on the later, that is correct, numbers were unsatisfied with this decision.

New York Post.

Will America invade Venezuela?

On January 3, the US bombed Venezuela, sent in special forces teams that successfully captured President Maduro, and announced that they would thenceforward, quote, run the country, a claim they later walked back.

Does this qualify as an invasion, in quotes?

Polymarket's resolution criteria defined invasion as, quote, a military offensive intended to establish control over any portion of Venezuela, end quote.

It didn't seem like the US was trying to establish control over Venezuelan territory, exactly, so they resolved no.

Traders who bet on yes were unsatisfied with this decision.

With one exception, these aren't outright oracle failures, they're honest cases of ambiguous rules.

Most of the links end with pleas for Polymarket to get better at clarifying rules.