Originally Posted by trips2win
It seems you contradict yourself here. Outliers exist everywhere (many of them post on this board...), but I guarantee a statistical model would show that 90-days is a significant sample for predicting future behavior. And I also guarantee that 90 days is much more indicative than 2 months, 1 month, 1 weekend.
Not necessarily. Consider qualifying based on one trip. This trip happens in one weekend, say. So this consitutes the activity for one weekend, one month, two months and 90 days. In this case, lengthening the period adds no more statistical significance. And this is consistent with many challenge scenarios.
Cheers.