In 2003, for example, there was Vahagn Mkrtchian, as well as Adibekian, who gave Kocharian victory in the first round (56% nationally). On another hand, there was the apparently opposition-ran Center for Election Technologies, Ashot Bleyan’s and Ruben Torosian’s outfits who would give Kocharian between 17-25% in Yerevan.

There was an interesting twist w/Aravot-Vox Populi, which would first show a rise in Kocharian’s popularity in Yerevan (from 24% in November 02 to 29% in late January 03), and then a fall to 22% just before the elections.

We’ll see if APR Groups follows a similar pattern.

If you know more, I think either an article for The Armenian Reporter, a blog post, or another comment would be appreciated. It would be really useful to finally understand how they function and if they can be trusted.

Gevorg Poghosian is always accused of making up his figures and always described as pro-government even though the U.S. seems happy to commission him. What’s your take on his polls?

According to Vox Populi group then contracted by Aravot and using the same methodology as APR Group now (and adjusted to subtotal) Robert had 24% and Stepan - 20% in early December in Yerevan. (That was even before endorsements from Vazgen and Raffi for Stepan.)

A conclusion is that Serge now is in a somewhat better place than Robert was in terms of popular support in 2002 and Levon is in a much worse spot than was Stepan.

In the first case, where Serge gets 30+% the percentages are counted from the subtotal of respondents who plan to vote for a candidate or are undecided but still plan to vote.

In the second case, the % are counted as share of all respondents including those who say they don’t plan to vote and those who are against all candidates (while in reality the views of these last two categories, comprising about 30% of all respondents won’t mathematically matter on election day).

If you adjust APR methodology to ASA methodology, you get 27% for Serge and 13% for Levon.

Another important point is that the pro-authority candidates (including Levon in 1996) did consistently worse in Yerevan than in the provinces. Add to that a month that had passed between the latest national ASA poll in late October and the latest APR poll in early December - the conclusion is that the two polls are not at all contradictory.