Here's a good way to do taste tests. Have one person mark the cups, like A B C or 1 2 3 or ! @ # etc. In threes.
He secretly pours two of the cups with Coke and one with Pepsi. Or vice versa. He makes as many triples as you like so everybody can have their own.
He sends trays of the marked and filled cups out to the tasters; he doesn't talk to them or watch them. They each say which of the three is different, and whether the different one is Coke or Pepsi.
If they can't tell which one is different, then they really can't tell them apart. If they can tell which one is different but they can't tell which is which that's better.
In my experience most people can't tell which one is different. They have definite opinions about which they like better, but they can't tell by taste.
After they find out that they can't tell them apart when the cups are not labeled, they still keep their definite opinions about which they like better, and they still think they can tell the difference -- after they see the label.
There are various tricks people claim they can use to tell which are which. Some people think they can tell by the size of the bubbles, or the number of the bubbles. Some people think they can tell a vanilla or citrus taste. Some people think they can tell by the amount of sweetness. In my experience, people who say this have consistently been surprised when they couldn't tell which one was different. They tended to doubt that one of them was in fact different. But they tended to doubt that they could tell which one all three were, either.