On 8/30/06, Steve Summit <scs(a)eskimo.com> wrote:
Greg Maxwell wrote:
It could exclude duplicate sampled hits, but then
it's not clear what
the heck it's measuring.. it's not measuring unique views, it's not
measuring pure views.
Sampled unique hits. What's wrong with that? (No worse than
sampled hits, it seems to me.)
You can't sample unique hits, at least not without first capturing all hits.
You could have unique sampled hits, but that doesn't map into any
metric anyone uses... and it will be misunderstood as sampled unique
hits which is another thing entirely.