The delayed choice quantum eraser is often explained as being more 'spooky' than it actually is.
The interference patterns only appear when you compare coincidences between the detectors. It's easy to see you could get an interference pattern after the fact by taking a bell curve (no interference) and subtracting out a sine curve (interference).
but how do you get the sine curve in the first place? because you're not subtracting a sine curve, you're subtracting non-coincidences, for which there is no reason they would be on the sine curve
