Party ID folks, party ID.
This has gotta be the 20th time where I've been forced to re-iterate my point about SUSA and party ID. You can't just throw SUSA polls out because their partisan spread looks off.
Okay I'll throw it out because it has too many young voters and the regional breakdowns are off.
In 2004, 17% of Virginia voters were under 30. Assuming that only 1/4 of the 30-44 demographic in Virginia was under 35 (probably generous to your argument), we're now up to exactly 25%. Which is, coincidentally, exactly the sample this poll has.
What about the regional break-down is a red flag to you? Keep in mind that the MoEs on those range from +/-7.2% to +/-8.3%.
