Does the 2018 Mississippi Special election really matter?

[Kyng]

Well, let's look at the demographic breakdown of both Mississippi and Alabama:

According to the 2010 US Census, Alabama was 68.6% White (67.0% non-Hispanic White), 26.2% Black, and 5.2% Other or Mixed. (3.9% were Hispanic or Latino of any race, so so I'm guessing about 2 of that 5.2% is Hispanics who don't self-identify as anything else, and 3.2% is Asians, Native Americans, Mixed, and so on)

Compare this to the CNN exit poll for the 2017 special election: 66% non-Hispanic White, 29% Black, 3% Hispanic, 2% Other. Essentially, blacks are overrepresented by roughly 3%, and the other racial groups are underrepresented by roughly 1% each.

Now, let's look at Mississippi. According to the 2010 US Census, this is 59.1% White (58.0% non-Hispanic White), 37.0% Black, 3.9% Other or Mixed (2.7% Hispanic or Latino of any race). If we apply similar shifts to those in Alabama, we'd end up with an electorate which is 57% non-Hispanic White, 40% Black, 2% Hispanic, 1% Others.

Now, in Alabama, Doug Jones got 30% of the White vote and 96% of the Black vote (the other groups were too small to show up in the exit poll). Barring a Roy Moore situation, I won't be as generous to the Democratic nominee in Mississippi: I'll give them 92.5% of the Black vote and 67% of Hispanics and Others. This alone would get the Democrat to (40 * 0.925) + (3 * 0.67) = 39% of the popular vote. The remaining 11% would need to come from non-Hispanic White voters - so, the Democrat would require (11/57 * 100) = 19.3% of the non-Hispanic White vote. Not the 30% that Doug Jones got in Alabama - but then, the GOP nominee isn't Roy Moore either.

So, with some generous assumptions - including a demographic turnout profile similar to that seen in Alabama - it'd be doable.

I'd still consider it to be Likely R: a big ask if the GOP candidate is even half-decent, but you never know.