The following is an analysis of what polls have been saying regarding tomorrows' Assembly Elections in Venezuela.
In the 2009 referendum to extend term limits, the results were 55% for Yes, 45% for No. (a margin of 9.73%. The polling firm IVAD predicted a 8.0% victory for the Yes. That's pretty accurate.
In the 2008 regional elections, Chavez' party won by 58%-41%. IVAD predicted a 13% victory for Chavez' party, which is fairly accurate, and even underestimating Chavez' lead. So that's fairly accurate.
In a defeated constitutional referendum in 2007, the anti-Chavez won by about 1-2%. IVAD had predicted a 4% margin anti-Chavez win. This turned out to be accurate.
For tomorrow's elections, IVAD predicts a 54%-46% lead by the pro-Chavez party.
Another historically accurate pollster is: GIS XXI, which, although being "clearly government-alligned," predicted a 11%-16% pro-chavez victory in 2009. the final result was "just under 13%". The final results were 10%-margin, so that's fairly accurate.
Opposition-alligned pollsters have been historically a joke. Read about their way-off predictions and refusal to release their methodology:
http://axisoflogic.com/artman/publish/Article_61200.shtmlMy initial advise: for analysis of Venezuelan elections, do not resort to American columnists or journalists from countries that hate Chavez. They will always exaggerate reality in favor of the opposition. Look at numbers. Look at polls. Serious polls.
Datanalisis is very bad, and Hinterlaces and Keller y Asociados are terribly bad, as you can see by reading their past predictions vs. actual results.