Quote from: Tommyhawk on February 15, 2012, 01:33 pmPersonally I believe the delivery rate should be increased by 1-2%. Scammers who say their packages have not arrived must be accounted for. We know they exist, we know they're frequent, and often there's no way of telling if they are lying or not. We can consider the scammers, the margin of error, in our whole percentage.There are basically four options for the Delivery Failure Rate (100 - DSR):1. Interception.2. Scammer Buyers pretending they didn't receive package.3. Lost mail (more likely than regular post due to Buyers goofing around with the address and name).4. Packages arriving late, such that the buyer reasonably could say the package didn't arrive and has no incentive to update feedback status. Similar to (2), but unintentional.The only way I can think of to reasonably separate those options from each other is to put electronic tracking devices with G.P.S capability. We also need a electronic device which detects the package being opened (yes, this is possible via photosensitive electronic devices) and sends us a timestamp to a busy open location online e.g. google groups/usenet etcTo prevent LEO using the new package profile, we'd need to send all packages simultaneously.Now, all of that is possible, and would allow us to work out exactly how many packages per 1000 get intercepted.The problem is a matter of funding, coordination and a significant level of technical expertise. Also the attention that such a sophisticated setup would attract. The keyword here is funding, we'd need ~200k to make that work. Product has to be real remember. Probably not worth it at this moment, until the interception rate results in losses above 400k a quarter. Right now it's simply more economical to lose the occasional package.