There's something significant about this poll, apparently all the undecideds in the poll are white. So, expect a huge Vitter jump on election day.
Also, while studying some of the 2014 polls (mostly the Kansas ones) and this year's Kentucky polls, I realized that the error in polling couldn't just be explained by undecideds completely going one way, it was also partially explained by the margin of error. Paul Davis was over 46% in many polls last year, but he ended up only getting 46%. Same for Conway in Kentucky (he was getting 45% in most polls, got 44%). If Louisiana follows a similar pattern, there is still a chance (a very outside one at that, though) that Vitter could win 51-49*. Most polls have Edwards at 52-53% with a margin of error at 3-4%. I don't think it will happen, I just won't be as shocked as some other people if he actually makes it uber close.
*There is actually something worse than the margin of error in some cases, saying your "probably" going to vote for the Democrat in a poll and then switching Republican at the last minute. May have contributed to Conway and Davis' fate.
It does seem very interesting how polls often seem to be incredibly biased towards the Democrats in traditionally Republican states compared to the actual results. There are a heck of a lot of examples of this, and it's getting kind of scary, TBH.
