Annually’s long enough for large-scale trends to dominate. I think we’d really have to be embedded in current events for our seven-day predictions to be much better than random, and closely following the press or news blogs is anything but fun—they deal in exaggerated significance, and filtering that out is exhausting.
A year is a long time, though. I think the ideal period might be something around quarterly.
Annually is not often enough to be fun. Why not make minor predictions seven days out every Sunday? You’d be able to score yourselves better.
Annually’s long enough for large-scale trends to dominate. I think we’d really have to be embedded in current events for our seven-day predictions to be much better than random, and closely following the press or news blogs is anything but fun—they deal in exaggerated significance, and filtering that out is exhausting.
A year is a long time, though. I think the ideal period might be something around quarterly.
Predictionbook seems better suited for this purpose. Also, I think longer-term predictions are more rationality-loaded.