BAGHDAD, Iraq, Feb. 3 - Iraqi and American election officials cautioned Thursday that there were several reasons that initial election returns could not reliably be used to project final results, beginning with the small number of polling stations involved.
Beyond indicating that the stations were somewhere in Baghdad and five southern provinces, officials also did not say exactly which neighborhoods were represented in the tally. And in a nation sharply divided into ethnic and religious enclaves, the early results could easily be skewed depending on where the voters lived.
One election official said that the selection was so complicated - or, perhaps, haphazard - that even within single polling centers, which generally had five or six individual voting stations, some stations may have been included in this first tally and others left out.
...
When asked why the commission had not done the arithmetic, one official asked rhetorically, "You mean, why haven't we made it easy for you to do an analysis that we consider unsound?"
http://www.nytimes.com/2005/02/04/international/middleeast/04vote.html