Issue with the 2004 Georgia Election Results

[Sorenroy]

So I was going through the 2004 election results for a project I'm doing and I noticed that Georgia seems to have a difference of a couple of votes between the Atlas (and Wikipedia) and the Georgia SOS.

The Atlas (and Wikipedia) lists the results as:
Bush — 1,914,254
Kerry — 1,366,149
Badnarik — 18,387

But the SOS office lists the results as:
Bush — 1,914,256
Kerry — 1,366,155
Badnarik — 18,387

Besides the fact that the SOS doesn't include Write-Ins (official or otherwise) which is frustrating, they also list an additional two votes for Bush and an additional six votes for Kerry. The differences come from three counties: Richmond, Walker, and Ware. The SOS finds an additional one vote for Bush and six votes for Kerry in Richmond, an additional one vote for Kerry in Walker, and an additional one vote for Bush and a loss of one vote for Kerry in Ware.

I would normally assume that this was an issue with Wikipedia and the Atlas and not the Georgia SOS, but the official FEC results match the Atlas and not the SOS. Does anyone know what these differences are about?

[MisterElection2001]

Maybe there was a discrepancy between the total reported by the SoS and the total of all the county reporting agencies? David corrects the vote totals if there is.