"2,508 adults, including 1,956 registered voters"
90 days from an election they should be doing better sampling than that.
Don’t look at me. If they included me in the sample they would have gotten “Somebody Else” as my preference.
I used to search their site and post the tricks pulled by Pew. I don’t bother anymore, but I remember them doing things like the question they actually asked was not the polling question they announced, or they took responses that belonged in the “no” category and added them into the “yes” category. They are very bogus.