The superdelegate count is not a poll. Polls are taken from a sample of larger target group. The margin of error is based on sample size vs. the size of the group being estimated. As the sample size increases, the margin of error decreases, and when the sample size is the entire group, the margin of error is zero. Your analysis using that 10% number is totally whack, the margin of error on the superdelegate “poll” is zero. Everyone who will be voting is part of the poll.
The biggest reason polls don’t match up to actual votes is because the sample doesn’t perfectly represent the whole group, not because people change their minds at the last minute.
If a committed Clinton superdelegate changes her mind and wants to support Bernie, she will publicly announce it when it happens. 300 of these people aren’t going to do that all at once, especially as Bernie is insulting them and the democratic party they represent. The superdelegates are actual people, people like Barbara Boxer, Jerry Brown, Harry Reid, Nancy Pelosi, Al Gore, Debbie Wasserman Schultz, Charlie Rangel, Chuck Schumer, Bill Clinton, etc. These are Democratic party heavyweights, many actively supporting Clinton (and getting more and more fed up with Bernie) as we speak. These are the superdelegates that Bernie’s supporters were stalking, harassing, and publicly smearing six weeks ago. They’re not changing their minds to support the delegate and vote loser with a worse chance of winning (you don’t believe this, but committed Clinton superdelegates do), and who’s being a douche. Although I bet Bernie will lose some more of his after Tuesday, he’s the only one who’s lost any committed superdelegates so far.
Hope all you want and vote your ass off for who you like best, but please don’t believe this crazy shit that you’re saying. You’ll just be embarrassed about it later. I’m embarrassed about it right now.