j reacts to "public policy polling"

"Public Policy Polling surveyed 1,276 likely voters on July 29th and 30th. The margin of error is +/-2.7%. 80% of participants, selected through a list based sample, responded via the phone, while 20% of respondents who did not have landlines conducted the survey over the internet through an opt-in internet panel."

you can throw them in the trash, too. i'm still waiting for a good firm to present itself.

with something like this, it just opens up the possibility that that 20% is being used to falsify the results. 1,000 people is a large enough sample this early in the race. there's no reason to bring this opt-in panel in other than to fuck with the results.

so, i don't want to naively accuse them of polluting the sample. no. they're doing it on purpose.

if they'd like to release just the real polling data separately, that would shut me up. don't count on it.