5 ms·
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 convincin
by oasisbob 16y 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.