4 ms·
The problem with this is that predictions should, ideally, be clearly specified, offer a percentage estimate (0-100% likelihood), and have some reliable method
by Emma_Goldman 8y ago
The problem with this is that predictions should, ideally, be clearly specified, offer a percentage estimate (0-100% likelihood), and have some reliable method of validating the outcome.
The Good Judgement Project does all these things: https://www.gjopen.com/ https://www.gjopen.com/
This website, on the other hand, seems to encourage none of these epistemic virtues. It's people making more or less precise verbal predictions on twitter and then assessing themselves.