While losing the capital, Pashinian acquired strong support in rural regions such as Shirak (65%). Many natives of this region became the first POWs in the 2020 Karabakh war. Although many these POWs
are still being tortured and killed in captivity, Pashinian negotiated release of a few Armenians, which raised his approval rate in the region. The region of Suinik, a de-facto active warzone, which lays on the pathways between the Azerbaijani exclave of Nakhichevan' and prevents direct communication between Baku and Istanbul, for a while did not allow Pashinian to come in considering him a traitor for the conditions of ceasefire agreement. Continuing instability and Azerbaijani attacks on Armenian border villages, explained by disagreements in demarcations, changed opinions in Suinik and its population voted for Civil Agreement to prevent further exacerbation.
But what about the Gallup polls that predicted 24.1% for the Civil Agreement and 23.8% for the Alliance "Armenia"? As in the US elections of 2014, polls were wrong, mobilization of rural population was not considered, and the result was unpredictable. However, situation in Armenia shared no similarities with the situation in the US. As it appeared, the Gallup polls were conducted not by the US Gallup Organization, but by Marketing Professional Group (MPG LLC), a member of the Gallup International, which uses the Gallup brand. The MPG is related to Kocharyan. Allegedly, the MPG "Gallup" polls were disinformation.