I said earlier that "In my experience about as many errors are introduced by scores being entered with the wrong declarer as those that are avoided by incorrect arithmetic or illegible handwriting", but I wouldn't have bothered quibbling with Pran had it not been such an extreme difference in perception. There are probably more that aren't noticed - one thing we've discovered is that the amount of information available on the web after events means that more errors are discovered that might well have been overlooked in the past. With Bridgemates, later players also often notice surprising scores and bring them to the attention of the dirctor, and most software also flags up potential errors of this sort.