The first # is rating, which is millions of viewers as determined by statistical analysis from the sample, (e.g., Nielsen). The second # is share, which is % of TV households on at the time -- also done by statistical analysis from a sample. A rating of 10 means 10 million viewers. A share of 10 means that 10% of TV households actually watching at the time were tuned to that show.
As an extreme example, if only one person in the country was watching ____________, that show would have a 0.0 rating. (Technically, 0.000001). However, if that person was the only person watching TV at the time, the share would be 100.0! (Assuming, of course, that this one person was part of the sample taken by Nielsen or whomever.)
Sorry, I had a brain fart. Rating isn't a direct represenation of any # of viewers. It's a %, just like share. However, rating is the % of all TV households. Share is the % of TV households using TV at that time. So, a rating of 10 means 10% of all TV households, regardless of whether they are watching TV or even at home. You can estimate # of viewers with rating, but it means nothing directly about # of viewers.
No shit! And that's against really tough competition. "Watching Ellie" (which isn't a lock for renewal, I know) got a 5.1 in its second week, for example. It lost to JAG, for cryin' out loud.