Most Nielsen data ties together with a program code, they often don't miss assigning the same program code to the different spellings. Local data arrives with BBTheory, BigBTheory, BBTheory087, etc. All of them will have the same program code. It is often up to the person pulling the raw data how to calculate the program averages. I am not sure who the one entity is in this case that is not including them in the average.
This is one of those things that strike me as so pathetic that I have a hard time believing it. Not that The Verge has the highest journalistic standards, but they’re non-click-baity enough that I’ll take them mostly at their word. And yet I was still reading the article to find out what the catch was. Nope, no catch; it’s just as sad as it sounds, with networks thinking Nielsen would <i>never</i> catch on to their clever ruse.<p>Hmm, what happens when Google catches someone trying to game the system...