Think of the implications of this social experiment.
Did you see the preview image?
You know, this one? Go ahead and study it.
How about a replay?
How about in slow mo?
Camera 2 in slow-mo?
Who would you pick in the lineup?
Let's see if you were right.
If you got it right, good for you. But let's be honest: You probably didn't. The vast majority of the audience didn't either.
Ordinary eyewitnesses are less like you and more like the people in the audience: completely unprepared to see something they really need to remember.
You saw the crime over and over. Full speed. Slow-mo. Freeze frame. Multiple camera angles. And you still probably didn't pick the right guy.
Yet we hold up eyewitness testimony like it's nearly infallible. It's not.