Quality
Posted by michael on March 22, 2001 at 15:08:15:

Yes, there will now be further judgement of the hit for significance. It's debatable whether or not that is something that should be included in the table, as I could be looked at as being subjective. It might just be bect to leave the raw probablity on the table and let everybody draw their own conclusions in their own ways.

Roger and I have agreed on how to grade the significance of hits, namely by the formula:
significance=100-probablity(%).

If a prediction has a 100% chance of occuring by random chance, then it is completely useless with a score of 0.

If a prediction has a 1.4% chance of occuring by random chance, then it is very usefull with a score of 98.6.

A miss receives a score of 0.

Predictors ability (initially)=total score/total predictions. This function may change as we look into the issues further.

Like I say, probablity might me the only item that belongs on the table as it is pretty much fact with very little subjectivity.

Michael


Follow Ups:
     ● subjectivity - Dennis  15:22:25 - 3/22/2001  (6292)  (0)