We do not understand the collapse of a wave packet by measurement and always ask the questions about the superluminal effects between the entangled pairs! Actually, this is a natural thing for the probability. The wave packet is just the "sample space", which contains all the possible results under a certain constraints. The measurement is just to pick up a sample with a certain method! It is not a collapse, it is a specific result of a probability event. If we want to construct the wave packet (the whole sample space) we should do enough measurements on it. For example, the wave packet of a micro-coin is \psi=0.5 up+0.5 down. Then measurement is just toss the coin and get the specific result, up or down! Then \psi will become up or down, and it is not a collapse! It is a very natural thing for probability events. Is it right?