tedster - 6:53 am on Sep 16, 2008 (gmt 0)

I've read through a couple times now, and it seems to me that the human evaluator team may be used more in the evaluation of low volume experimental SERPs than I was originally thinking.