But then the original prediction would still count against your overall score when trying to achieve the statistically significant level.

But then I guess our goals are different.

One view is for the sake of evaluation that limits prediction posting requirements.

The other view is for the sake of giving people the best warning that can be given based on any information available at the time and any other information that may come available later on.

So it would appear that one has to suffer in order to achieve the other except that in order to achieve the other, you need to achieve the prior. Is that something like Catch22?

Dennis