ioi speaks out about ergh "VGC analysts"

[Torillian]

While I understand where you're coming from Brett, I think the discrepancy between what readers think is accurate and what you'd call accurate is partially on how the numbers are reported. Is it common in statistics to report numbers to such an apparently high precision even though you know the error to be higher than that? In a scientific journal the standard is to report the estimated error in the results and to round the number to the first significant digit which with a 5% error would be probably be the third digit so 282,965 would become 283,000. Is there any particularly issue with reporting the numbers this way with the estimated error included versus what you do now?