06-05-2004, 03:45 PM
If you have a statement like that, prove it. Show us the results of 1,000+ controlled MF runs, recording the quality of each dropped item (including failed uniques/sets). Show that there's a decrease in quality over time. Then do your "trick" and show another 1,000+ results, showing that this improved your chances. If it really has an effect, we should be able to see it in the data.
(I'm only guessing with 1,000+ value. I'm not sure just how many samples we'd need for a statistically conclusive result).
(I'm only guessing with 1,000+ value. I'm not sure just how many samples we'd need for a statistically conclusive result).