Let's just assume you're right, and this experiment is true a majority of the time... wouldn't another possible explanation be that that's a perfectly fair representation of things? Both sides aren't always equal. Weighing the coverage of both sides to be equal would be misleading.