← Back to context

Comment by discreteevent

3 months ago

Is this really a rational strategy for something whose nature is to be right most of the time and then spectacularly wrong a much lesser amount of the time?