AA,
That's an excellent question and if the particles were, say, "red" or "blue" then you would be correct. What makes it intresting is that unlike colors which don't change, spins can change. Please consider the following;
Imagine Each particle has a spin which can take one of two values, let's say +1 or -1 (for electrons it's actually +/- 1/2 but it makes no real difference).
Now, here's the twist, the spin is defined with respect to some coordinate system, let's say x, y, z. So you can measure the z-component of spin, or the x-componet or the y-component, whatever you like. Think of it like a ball spinning around a particular direction in space (again, this is a simplification but it's ok for the discussion). Now a funny thing about QM is that you can't know all three spins at once, you can only know one of them. So, let's say you measure the z-component and get +1. What value would you get had you measured the x-component? You have no idea execpt that you would get either + or - 1. It can eb either but you don't know which you will get until you mak eteh measurement. 50% of the time you get +1 and 50% -1. It's the same for all directions.
Now imagine you make 2 measurements and get
z-component: +1
x-componnet: -1
now you measure the z-component again, what do you get? You might think it would still be +1 - just like if you "measure" a ball's color to be red, then weigh the ball, you don't expect it's color to change to green. But when you measure the z-component again you get + or - 1 but which one you don't know until you've measured it. So you could get;
z-component: +1
x-component: -1
z-component: -1 So spin isn't like color that is fixed. The actual spin you measure isn't known until you measure it. Now, if you **always** measure the z-component you always get the same result as the first time you measured it, but if you measure x or y in between then that disrupts the z-measurement.
So this demonstrates that particles don't have known spins until they are measured. You only need one particle to prove this. Now consider two particles;
Imagine that you and AlanF go to opposite sides of town but have secretly agreed ahead of time that you will measure the spins of particles that pass you by, but that each time you will measure a different spin component. You decide on the order z,y,x,x,z etc. Also you decide that you will go first.
Obviously the particles don't know what you agreed. They have no idea whihc component you will measure first and, if you believe the above, they can't have pre-assigned values for all three components.
So, two particles arrive, one at you and one at alanf. Now you know that whichever spin component you measure you will get +/-1 but you also know that you can't know all three spin components at the same time. So it's not like the particles have pre-assigned values for x, y, and z components - that is, it's not like they have special colors.
Now, being a sneaky bastard you decide to measure the x-component first and not the z-component as you agreed and you get +1. AlanF being equally sneaky also decides to cheat and has also by luck, measured the x-component. He gets -1. Had he measured teh x-component as agreed he'd have got +/-1.
Now imagine this, the next two particles coming flying out and unbeknownst to you someone between you and the origin measures the z component of spin and gets -1 for your particle. Then you measure the x-component and get +1. Then AlanF measures the x component - he can now get +/-1 so he might also get +1. The entanglement was broken by the intruder's measurement. When you compare notes you can tell in thi scase that someone broke the entanglement. Of course, if AlanF had got -1 (which he would 50% of the time) you couldn't have told. So you need to look at the statistics not at any single measurement to detect the intruder.
With the ball analogy if an intruder had a glimspe of yoru ball then that wouldn't have altered it's color.
This is a tricky subject and I hope this helps a bit
