Very interesting video, and I really like the practical demonstration of how the same "dataset" can mean something completely different depending on what you expect from it.
Its a nice way to show that even though you might consider yourself "completely unbiased" and taking the scientific approach, you might get the wrong result anyways.