|
|
|
|
|
|
by oasisbob
5794 days ago
|
|
|
If your numbers are large enough, no. As long as the selection isn't biased, it doesn't matter. Another commenter on this topic has a good post about convincing management of the validity of sampled metrics, I'd read it then tip your chair back and stare at the ceiling for a while. |
|
|