Aside from that, a N=1 does not say anything about how a ROW does or does not increase probability on a rare item. With extremely rare items, an extremely large sample size is needed to measure effects irregardless of what the exact mechanism is by which a rare drop is calculated. You'd need to try getting ~50 rare drops with and without ROW, for example. Then you'd be seeing some effects. It perfectly explains it. The drops we as players consider rare are in the "bonus drops" ROW affects the odds of you getting a rare drop from the normal drop tables. Bonus drops are unaffected, therefore ROW is of no use in getting the drops we as players class as rare. But something like, lets say, a whip doesn't fall in the bonus category right? And then there's still the point that you can't compare probabilities without a decent sample size. Although it will not work on bonus items irregardless of a ROW, you can't just simply say something about probability in a situation stated in the first post. Although your post does indeed explain the mechanism, I'm saying it's more than just the mechanism. How you measure it is important as well.