Yeah, I imagine it's pretty annoying especially for those in the running of winning something. I bet it could be quite easily done to measure the effect of the merit awarding, since there is definitely a reoccurrence of certain users since they have been giving merit, which didn't seem to be in the running at the very start. Although, that could just be exposure, and natural variation, so it probably wouldn't prove much at all. Although, I do believe the OP has the ability to not include anyone's votes at their discretion.