34 million are Opt-Outs.
22 million are On-Demanders.
I wish companies that send out announcements and nicely formatted summaries would also attach raw data, because the way these segments are defined here seems kind of arbitrary. For the Opt-Outs, how many people don't own a TV at all compared to last year? Is not having watched TV in the past week indicative of their long-term behavior? For On-Demanders, how many of those who watch less live TV are streaming? How many of those who stream watch more live TV as a result?
There must be some people who watch more TV for the overall viewership time to have gone up, right?