Judging this by the method used (counts of uses of top 5 hashtags associated with the conflict) is ludicrously bad as a methodology, because, aside from not looking at sentiment, its prone to being radically wrong if one side is more consistent in hashtag use than the other.