Nate Silver noting the polling in Alabama is close (within the margin of error). That means Jones could have a narrow win, or Moore could have a blowout.

He notes the same happened with Ralph Northam in Virginia (he was up by 3, he won by 9).

fivethirtyeight.com

He also notes the margin of error in Senate polling since 1998 (in normal elections) is 13 points. In a special election that is probably worse.

Also figuring in is how many people would not want to admit they are voting for an alleged child molester, versus how many say “to heck with this” and choose to stay home and not vote (thus benefiting Jones).

Dave Weigel (who is not a pollster) notes the following to make it interesting:

Before everything turns to #MNSen when Franken speaks at 11:45 AM, I put together a preview of the #ALSen race for this week’s Crystal Ball. There are many maps & I consider just how a Doug Jones win might look. Here is that map: https://t.co/n2ASO8IkgC pic.twitter.com/Fo4J3h87AY

— Geoffrey Skelley (@geoffreyvs) December 7, 2017

The best possible Jones map has him winning 67% of Jefferson County. Which is actually doable.

How much has Alabama changed? In 1976, when Carter became the last Dem to carry Alabama in a prez race, he won 60 of 67 counties. He lost Jefferson County. https://t.co/pq2zLhQapZ

— Dave Weigel (@daveweigel) December 7, 2017