It’s not about errors, as HD’s have various methods (ECC) to ensure such things are not the case. I was speaking of ensuring that 100% of the old data was wiped, such that it could not “bleed” through to a new install. Most think all they have to do is high level format and the data is gone, and that’s not true. Only a low level format can assure such things, unless you’re not really worried about security, but rather just ensuring the old install is in fact totally gone, in which case if you boot to an OS on an optical drive, which of course is read only, without a swap file, format the HD to FAT32, (there are partition size restrictions, I keep my OS partition rather small, just enough for the OS, my swap file is in the first partition of another physical drive, for increased effective speed), so you do not have any of the hidden NTFS files, like the USN Journal, etc, leftover, then fill up the drive with useless data, I use 1 gig vob files, and smaller ones at the end, then delete all and do it again, starting with the small ones to ensure that the 4k clusters don’t line up from the first to the second write, then reformat with NTFS I assume, then install the OS. This way you are guaranteed that the old install is gone, and relatively sure of security, at least of all the non mapped out sectors, only low level formatting can deal with those, but they can’t “bleed” through to an new install, if that’s all you’re worried about. I hope that’s clearer