pull down to refresh
Temperature in Paris wasn't above 35° or whatever the cutoff was. This song wasn't top on Spotify (other than via some dude's bot attack)
The temperature was above that in the one part of Paris that mattered, which was specified in the resolution criteria, and apparently that song was the top of Spotify’s poorly designed chart.
No it wasn't. You can't honestly argue/mean that. Some dude with a heater made that thermometer say that.
Wasn't real, wasn't capturing what "thermometer" means
My point, that I do mean, is that resolution criteria need to account for things like that and be robust against attempts at rigging outcomes.
Participants also need to learn to think critically about non-obvious ways the criteria might be met.
Everyone’s still pretty new to this stuff and mostly they aren’t good at it yet.
Game of whack-a-mole, can't win. Goodhart's law, instrument becoming a measure and fails to be an instrument
That doesn’t mean there aren’t better and worse approaches.
How are these things false? People need to read and think about the decision criteria more carefully.