The 2014 primary season ends (Sept. 9: DE, MA, NH, NY, RI)

[cinyc]

The Atlanta Journal-Constitution has a map of the GA-Republican Senate results here.  The regional variations are stark.  Perdue is running first or second virtually everywhere - so he should almost certainly make the runoff.

[cinyc]

^ Yeah, per AJC, Gwinnett is still not reporting, and we need to see the results there to determine if Atlanta is enough for Handel.

14/156 of Gwinnett is reporting.  Handel leads Purdue 34-31, with Kingston pulling 16.  I think Kingston is doing just well enough in Metro Atlanta to keep him in second.  We'll see.

[cinyc]

Kingston might be close to maxing out, percentage-wise. I think he still has the edge above Handel, but it could be very close, and Handel could still snatch second place.

Kingston really cleaned up in South Georgia.  He's leading in some counties with 70% of the vote.  Some of those counties are currently only showing absentee/early votes.  Granted, those counties are smaller than the Metro Atlanta big 4, but they greatly increase Kingston's overall margin. 

I haven't done any sophisticated math, but Kingston should have this, likely taking second.  Handel's margin over Kingston in the Metro Atlanta counties simply isn't large enough.

[cinyc]

Covington County, next to McDaniel's Jones County is out.  If it goes as much for McDaniel as neighboring counties and the Hinds County precincts are mainly African-American, it could put him above the mark for a runoff.

[cinyc]

93% now in MS, McDaniel leading 49.7 to 48.7. Most of the counties with only partial results are going for Cochran, but a few are going for McDaniel. Still nothing in from Holmes or Covington though.

Holmes County probably doesn't have many Republican primary votes in it.  Its population is around 19,000, but it's about 85% black.  Covington should have mote Republican votes.  Its population is also around 19,000, but it is 65% white.  It neighbors McDaniel's best county, Jones County.

[cinyc]

McDaniel down to 49.4%. Needs to gain about 3000 to avoid a runoff.

That was primarily a dump from Hinds County (Jackson).  The precincts weren't empty.  A runoff looks likely, barring a really strong pro-McDaniel turnout in Covington.

[cinyc]

The only way to avoid a runoff now is if Covington County nets 2500-3000 for McDaniel and McDaniel wins big in the remaining Rankin precincts. Not impossible, but not likely.

I don't think McDaniel's Senate District goes into Covington County.  McDaniel's margin of victory there might not be as high as it would be if it were.

[cinyc]

Still got Holmes and 8 precincts out in Warren, 4 in Hinds. How many votes does Covington have?

Covington County had about 2,000 votes in the 2012 Republican Presidential Primaries, if I've done the math right.

[cinyc]

Holmes poll workers are saying that they will keep working all night if necessary to report the results.

I doubt it will affect the results much.  There were only 553 votes there in the 2012 Republican Presidential Primaries.  Even if there are 1,000 votes this time, that's not enough for either person to avoid a runoff.

[cinyc]

It probably won't take to long for other outlets to predict a runoff in Mississippi. 

Probably not.  There are likely at least 6,000 votes out from Rankin County - and perhaps as many as 9,000, if turnout was higher than 2012 - without any way to know which part of the county is out.

[cinyc]

From the current AP results, if there are 6,000-10,000 votes outstanding, McDaniel would need to win 65.6%-61.0% of them to avoid a runoff.   Not likely, based on the partial county results where the vote is out, but I doubt the AP will call a runoff without seeing more Rankin County results, in any event.

[cinyc]

About 2,500 votes just came in from Rankin County.  They went 50.5%-48.6% Cochran.  Not a big enough Cochran margin to change the statewide margin - 49.6%-48.8% McDaniel.  Rankin is now 84% in (8 precincts left to count).

All of Holmes County remains out, along with 4 precincts in Hinds County and 2 precincts in tiny Jefferson County.  Jefferson County is about 85% black, and has only cast 219 Republican primary votes so far.

McDaniel would need to win 87.5%-72.5% of the remaining expected 3,000-5,000 votes to avoid a runoff.  Impossible.

What are Mississippi's rules for counting absentees?  Are they included in tonight's tallies?

[cinyc]

Holmes County, Mississippi reported.  Cochran won it 58.4%-41.1%.  But there were only 625 votes cast in the Republican primary.  The statewide margin remains the same.

Only some precincts in Hinds (4 precincts), Jefferson (2) and Rankin (8) remain outstanding.

[cinyc]

5 more Rankin County precincts reported, casting 2,115 Republican votes.  Cochran won the precincts 55.3%-43.7%, putting him back in the lead in Rankin County by .4%.  Cochran picked up a (rounded) tenth in the statewide tally.  It's now 49.6%-48.9% McDaniel. 

It's now almost mathematically impossible for McDaniel to avoid a runoff, assuming there are less than 2,650 votes outstanding.

Rankin has 3 precincts, Hinds 4 and Jefferson 2 precincts outstanding.  The Jefferson precincts are tiny.  The makeup of Hinds' precincts is unknown and variable.  Based on 2012 turnout, Rankin probably has about 2,000 votes left to count.

[cinyc]

Also, to any people who know Mississippi: What the hell happened in Wilkinson County? It's the only county Childers lost, and not only that, he's in last place there. And the guy in last place overall leads there. WTF?

Good catch.  There has to be an error there.  The third Republican, who is receiving 1.6% of the vote statewide, supposedly received 25% of the vote in Wilkinson County.  Also, turnout in the Republican primary is supposedly running 50% of 2012, which is unheard of in most other counties, where turnout equals or exceeds 2012.

Luckily, the county is tiny (286 Republican voters in 2012) and the error won't affect the totals much.

[cinyc]

Cochran seems to be back up by a very slim margin (less than 100 votes).

Where?  In Rankin County?  Not statewide.

[cinyc]

Cochran seems to be back up by a very slim margin (less than 100 votes).

Where?  In Rankin County?  Not statewide.

I might have misinterpreted something in that case.

EDIT: Cochran + 69, according to Austin Barbour. This is after updates in Hinds and Rankin. I saw someone post it on RRH.

I'll believe it when the AP posts it.  The Hinds County precincts would have had to be very large and very pro-Cochran for that to be true.  And there are still two small Jefferson County precincts to be counted, which probably tilt Cochran to give him tens of additional votes.

Rankin County is probably done counting.  They posted a precinct-level report on their website.  Unfortunately, best I can tell, it doesn't include a county-wide total and compiling it using Excel would be a pain due to the format they use to report results.

[cinyc]

Based on a quick, lazy Excel analysis of the Rankin County precinct results, Cochran won the county 50.0%-49.0%, 11900-11659-249.   Cochran won the last three precincts' 2,548 voters 52.5%-46.6%.  

When updated to show full Rankin results, assuming my math is correct, McDainiel should lead 153,030-151,052-4,771.  Cochran would need to net about 2,050 Hinds County votes to be up by 69.  I guess it's theoretically possible - if there are 6,150 votes outstanding and the current county margin holds, or if there are 3,000 votes outstanding and Cochran won about 84% of them, Cochran would be ahead.  Who knows if that's what actually happened, though.

 

[cinyc]

With 4 Hinds and 1 (supposed) Rankin precinct left to count, McDaniel leads 49.5%-48.9%, per the AP tally.  

Based on the precinct-level data provided by Rankin County, the only outstanding Rankin precinct isn't one of the 50 real physical precincts - it's precinct 51 -  PVP Purged Voter Precinct - likely affidavits.

Per media reports on Twitter, the only ballot box yet to arrive at Hinds County election HQ is from Wildwood Baptist Church in Clinton.  Clinton is a suburb of Jackson with a population of about 25,000 that's 60% white.  It's not clear how big that particular precinct is though.  I can't find 2012 precinct-level results.  It's also unclear where the other 3 outstanding Hinds precincts are.

The Cochran camp is claiming on Twitter that Cochran will lead by 75, based on their Hinds numbers.

There are also an undetermined number of paper ballots that need to be counted in Harrison County (Biloxi).  Republican officials delivered the wrong boxes to 13 precincts.  While the error was caught, ballots can only be counted by machine if they are delivered to the correct precinct.  It is unclear how many erroneous ballots were cast - though the county is already at 134% of 2012 Republican primary turnout, so it might not be many.

Harrison County went to Cochran and Taylor.  This gives Taylor hope that he can force a runoff in MS-04.  He needs to net about 800 votes to do so.

[cinyc]

Rankin's now all-in.  McDaniel leads by 1,736.  The 848 votes in the last Rankin precinct (whatever it was) went 58%-40% Cochran.

Updated results from other counties also seem to be trickling into the AP tally - perhaps affidavits.

[cinyc]

AP gives Palazzo the check.

There were only 206 uncounted ballots from 8 supposed bad precincts.  Not enough to put Palazzo under 50%.

[cinyc]

Vote changes in Carroll, Hinds, Itawamba and Simpson counties shrink McDaniel's lead to 1,579. 

There are still supposedly 4 Hinds precincts out.  If one of them is Clinton 2, it had  2,620 registered voters on general election day 2013 and voted overwhelmingly Republican in local races.

[cinyc]

Vote changes in Lowndes and Warren counties shrink McDaniel's lead to 1,454.  I'm not sure what these votes are, but they have tended to be very Cochran heavy - in some counties, almost 70-30.

[cinyc]

Forrest, Madison and Oktibbeha counties with updated numbers shrinks McDaniel's lead to 1,362.  

If there are 3,000 votes outstanding, Cochran needs about 73% of them to pull into the lead.  Definitely possible, since Hinds was one of his best counties and whatever these new numbers are, they are typically coming in more pro-Cochran than the county average.

[cinyc]

Forrest, Madison and Oktibbeha counties with updated numbers shrinks McDaniel's lead to 1,362.  

If there are 3,000 votes outstanding, Cochran needs about 73% of them to pull into the lead.  Definitely possible, since Hinds was one of his best counties and whatever these new numbers are, they are typically coming in more pro-Cochran than the county average.

So, Cochran could get to 50% of the vote still?

Probably not.  But Cochran can pull ahead of McDaniel in the overall tally.